Skip to main content

Full text of "Proceedings : International Workshop on Subalpine Stone Pines and Their Environment: the Status of Our Knowledge, St. Moritz, Switzerland, September 5-11, 1992"

See other formats


Historic,  Archive  Document 

Do  not  assume  content  reflects  current 
scientific  knowledge,  policies,  or  practices. 


United^tes 
Department 
Agriculture 


Forest  Service 

Intermountain 
Research  Station 

General  Technical 
Report  INT-GTR-309 


July  1994 


Proceedings — International 
Workshop  on  Subalpine 
Stone  Pines  and  Their 
Environment:  the  Status 
of  Our  Knowledge 


n 


:3 


Statements  by  contributors  from  outside  the  U.S.  Department 
of  Agriculture  are  solely  the  views  of  the  authors  and  do  not  re- 
flect policies  or  opinions  of  the  Department. 

The  use  of  trade  or  firm  names  in  this  publication  is  for  reader 
information  and  does  not  imply  endorsement  of  any  product  or 
service  by  the  U.S.  Department  of  Agriculture  or  other  organi- 
zations represented  here. 


Intermountain  Research  Station 
324  25th  Street 
Ogden,  UT  84401 


Proceedings — International 
Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment: 

the  Status  of  Our  Knowledge 

St.  Moritz,  Switzerland,  September  5-11, 1992 


Compilers: 

Wyman  C.  Schmidt 
Intermountain  Research  Station 
Forest  Service 

U.S.  Department  of  Agriculture 

Friedrich-Karl  Holtmeier 
Department  of  Geography 
University  of  Munster 
Germany 


Workshop  Sponsors: 

intermountain  Research  Station 
Westfalische  Wilhelms-Universitat, 

Landscape  Ecology  Division,  Germany 
Swiss  Federal  Institute  for  Forest,  Snow, 

and  Landscape  Research 
Karl  and  Sophie  Binding  Foundation, 

Switzerland 


FOREWORD 

This  proceedings  is  a  product  of  the  first  comprehensive 
and  international  examination  of  stone  pine  species  of  the 
world.  It  reports  the  papers  that  were  presented  at  the  Inter- 
national Workshop  on  "Subalpine  Stone  Pines  and  Their 
Environment — The  Status  of  Our  Knowledge"  held 
September  5-12, 1992,  at  St.  Moritz,  Switzerland. 

It  is  generally  recognized  that  there  are  five  subalpine 
stone  pines  in  the  world— all  in  the  Northern  Hemisphere- 
one  in  North  America  and  four  in  Europe  and  Asia.  Ail  share 
many  of  the  same  characteristics  and  occupy  similar  ecologi- 
cal niches.  Stone  pines  are  defined  as  those  species  of  Pinus 
that  have  five-needled  fascicles,  have  wingless  seeds,  and 
have  cones  that  remain  closed  at  maturity. 

How  did  this  international  workshop  come  about?  Comple- 
tion of  a  whitebark  pine  workshop  in  1987  and  a  major  sym- 
posium in  1989  in  Bozeman,  MT,  U.S.A.,  sparked  interest  in 
high-elevation  ecosystems  of  western  North  America.  One 
section  of  this  popular  symposium  specifically  addressed  high- 
elevation  ecosystems  of  the  world.  Information  presented 
there  hinted  that  we  could  learn  much  about  whitebark  pine 
ecosystem  analogs  from  other  parts  of  the  world  that  have 
similar  physical  and  biological  features.  In  addition  many  of 
the  Eurasian  forests  have  a  long  history  of  human  influence. 
The  logical  place  to  look  for  this  type  of  information  was  the 
world's  forest  areas  that  supported  subalpine  stone  pines. 
A  cursory  examination  of  the  five  commonly  recognized  sub- 
alpine stone  pines  indicated  they  had  many  similarities. 

With  this  as  a  background,  a  small  group  of  western  U.S. 
researchers  and  managers  held  a  planning  session  with 
Dr.  Holtmeier  from  Muenster,  Germany,  in  Missoula,  MT,  in 
early  1990.  This  planning  spawned  the  notion  that  it  was 
time  that  we  started  doing  a  better  job  of  collaborating  be- 
tween colleagues  in  Europe,  Asia,  and  North  America.  The 
logical  starting  point  was  an  international  workshop  that  ad- 
dressed those  items  that  held  common  interest  for  European, 
Asian,  and  North  American  specialists.  So,  the  international 
planning  transmission  was  put  into  gear  and  2V2  years  later  a 
workshop  was  held  in  St.  Moritz,  Switzerland. 

The  primary  objectives  of  the  Stone  Pine  Workshop  were: 

1 .  To  improve  international  cooperation  and  collaboration 
between  scientists. 

2.  To  exchange  research  findings. 

3.  To  determine  knowledge  gaps  in  the  ecology  and  man- 
agement of  subalpine  stone  pines  of  the  world. 

All  five  subalpine  stone  pines  were  covered  by  authors  of 
48  papers  from  Europe,  Asia,  and  North  America.  They  were: 


Species 

Swiss  stone  pine 
{Pinus  cembra) 

Siberian  stone  pine 
{Pinus  sibirica) 


Natural  range 

European  Alps,  Carpathian 
Mountains 


Siberian  Russia  and  Northern 
Mongolia  (95  percent  in 
Russia,  5  percent  in  Mongolia) 

Japanese  stone  pine     Japan,  Korea,  Eastern  Siberia 
{Pinus  pumila) 

Korean  stone  pine        Northeastern  China,  North  Korea, 
{Pinus  koraiensis)       Honshu  Japan,  southeastern 

Russia 


Whitebark  pine 
{Pinus  albicaulis) 


Western  North  America 


The  1 2  countries  represented  and  the  number  of  papers 
from  each  country  were:  Russia  (9),  China  (1),  Japan 
(4),  Switzerland  (5),  Austria  (4),  Italy  (3),  France  (1),  Yugosla- 
via (1),  Czechoslovakia  (1),  Romania  (1),  Germany  (3),  and 
United  States  (15).  The  main  subjects  covered  were: 

•  Evolution  and  taxonomy. 

•  Basic  ecology  of  stone  pine  species. 

•  Growth  characteristics. 

•  Influence  of  environmental  factors. 

•  Regeneration. 

•  Importance  to  wildlife. 

•  Forest  structure  and  dynamics. 

•  Forest  management. 

•  Research  needs. 

Stone  pines  are  critical  forest  components  on  vast  areas. 
Pinus  sibirica  alone  covers  40  million  hectares,  mainly  in 
Siberia,  and  is  a  key  component  of  the  economy  and  biology. 
P.  l<oraiensis  accounts  for  one-fourth  of  the  timber  production 
in  China.  P.  albicaulis  is  key  to  the  survival  of  wildlife  such  as 
the  grizzly  bear  and  is  a  protector  of  high-elevation  water- 
sheds and  visual  resources.  P.  cembra  plays  a  key  role  in 
avalanche  protection,  wood  production,  and  recreation  in 
Europe.  P.  pumila  is  a  key  high-elevation  species  for  stabili- 
zation of  mountain  slopes  in  Japan  and  Kamchatka.  These 
species  have  many  things  in  common,  likely  starting  with  a 
common  ancestor,  probably  P.  sibirica.  They  also  share 
common  methods  of  seed  dispersal  with  the  North  American 
and  Eurasian  nutcrackers.  Natural  selection  processes  have 
resulted  in  a  real  mutualism  between  subalpine  stone  pines 
and  the  nutcrackers. 

This  workshop  was  sponsored  by  the  Intermountain 
Research  Station,  USDA  Forest  Service,  U.S.A.;  the 
Westfalische  Wilhelms-Universitat,  Landscape  Ecology  Divi- 
sion, Germany;  the  Swiss  Federal  Institute  for  Forest,  Snow, 
and  Landscape  Research,  Switzerland;  and  the  Karl  and 
Sophie  Binding  Foundation,  Switzerland. 

We  want  to  acknowledge  the  people  and  the  organizations 
that  supported  them  for  their  enthusiastic  and  dedicated  ef- 
forts in  the  planning  and  conduct  of  this  international  work- 
shop. Three  field  tours  were  conducted  by  the  hosts  of  the 
workshop.  The  first  tour  entitled  "Engadine  Cembra  Pine- 
Larch  Forest,"  was  in  a  cembra  pine-larch  forest  overlooking 
St.  Moritz,  a  heavily  used  recreation  area  in  the  Engadine. 
The  primary  subjects  were  tree  regeneration  and  succession 
in  avalanche  tracks  and,  at  the  upper  timberline,  squirrel  and 
nutcracker  interactions,  human  historical  influence,  grazing 
and  game,  and  the  pressures  of  modern  tourism  on  the  for- 
ests and  landscape. 

The  second  tour,  entitled  "Ecology  and  Technique  of  High- 
Altitude  Afforestation,"  dealt  primarily  with  management  of 
avalanche-prone  areas.  It  was  conducted  in  the  Stillberg  Ex- 
perimental Area  near  timberline  in  the  Dischma  Valley  near 
Davos.  It  demonstrated  the  use  and  design  of  physical  barri- 
ers for  avalanche  control  and  reforestation  methods  with 
cembra  pine,  larch,  and  mugo  pine  for  long-term  reduction  of 
avalanche  problems. 

The  third  tour,  entitled  "Subalpine  and  Alpine  Environ- 
ments," emphasized  vegetation  succession,  particularly  of 
cembra  pine  and  larch  following  the  gradual  retreat  of  the 
Morteratsch  Glacier.  Also  included  in  this  tour  was  a  visit  to 
the  glacier  areas  at  Diavolezza  on  the  Bernina  Pass  where 
recreation  values  are  paramount.  From  the  aerial  tram  to  the 


mountain  station  (2,973  m)  we  viewed  the  strong  contrasts  of 
vegetation  and  glaciers  on  the  north  slopes  of  the  Bernina 
Pass  area. 

For  all  the  hosts  of  the  enlightening  field  tours  and  arrange- 
ments at  St.  fvloritz  we  owe  a  special  thanks.  Those  included 
Drs.  W.  Schonenberger,  R.  Hasler,  J.  Senn,  R.  Lassig, 
F.-K.  Holtmeier,  and  Herr  U.  Wasem — also  Engadine 
Foresters  J.  Altman,  I.  A.  Bisaz,  and  G.  Bott.  We  also  thank 
Prof.  Dr.  H.  Steinlin  from  the  Binding-Siftung  who  attended 
the  workshop  as  representative  of  the  Binding  Foundation. 
All  of  these  people  and  their  staffs  were  wonderful  hosts  who 
contributed  so  much  to  the  success  of  this  first  international 
conference  on  stone  pines. 

A  special  thanks  goes  to  Kathy  McDonald,  Intermountain 
Research  Station,  U.S.A.,  who  processed  all  the  manuscripts. 

The  organizing  committee  was  composed  of  the  following 
people: 

Dr.  W.  Schonenberger 
Dr.  R.  Hasler 
Dr.  J.  Senn 
Swiss  Federal  Institute  for 
Forest,  Snow  and  Landscape  Research 
Zurcher  Str.  1 1 1 
CH-8903  Birmensdorf 
Switzerland 

Prof.  Dr.  F.-K.  Holtmeier 

University  of  f^unster 
Department  of  Geography 
Landscape  Ecology  Division 
Robert-Koch-Str.  26 
W-4400  Munster 
Germany 

Dr.  Wyman  C.  Schmidt 
USDA  Forest  Service 
Intermountain  Research  Station 
Forestry  Sciences  Laboratory 
Montana  State  University 
Bozeman,  MT,  59717-0278,  U.S.A. 


Drs.  Holtmeier  and  Schmidt  served  as  technical  evaluators 
of  all  the  papers. 

As  a  consequence  of  this  international  workshop,  a  resolu- 
tion was  passed  by  the  scientists  in  attendance  that  aimed  to- 
ward greater  coordination  and  sharing  of  research  findings. 
The  resolution  was: 

1 .  To  organize  an  International  Association  for  Stone  Pine 
Research. 

2.  To  form  an  international  organizing  committee  by  elec- 
tion that  could  probably  meet  in  Tomsk,  Russia,  in  1993  (or 
later)  to  (a)  establish  the  objectives  of  the  Association;  and 
(b)  establish  the  organizational  structure  of  the  Association. 

3.  The  organizing  committee  to  develop  the  preliminary 
plan  for  the  next  International  Stone  Pine  Workshop. 

4.  To  hold  the  next  International  Stone  Pine  Workshop  at 
Tomsk,  Russia,  in  1995  (or  later). 

The  world  faces  many  environmental  and  resource  chal- 
lenges. Many  of  these  challenges  are  in  temperate  forests  of 
the  Northern  Hemisphere.  In  these,  stone  pines  are  a  signifi- 
cant forest  component  and  one  in  which  our  knowledge  base 
is  certainly  not  complete.  Sharing  the  knowledge  developed 
from  similar  ecosystems  and  building  on  that  solid  base  is 
key  to  a  better  understanding  of  how  our  subalpine  forests 
function  ecologically.  We  can  learn  much  from  each  other. 
The  international  workshop  that  generated  this  proceedings  is 
a  good  first  step  in  that  direction.  Let  us  do  what  we  can  to- 
gether to  keep  the  momentum. 


Wyman  C,  Schmidt  F.-Karl  Holtmeier 


VORWORT 


Bei  den  Arten  handelte  es  sich  um: 


Dieser  Band  ist  eine  erste  umfassende  internationale  Studie 
iiber  die  subalpinen  Steinkiefern  der  Erde.  Er  enthalt  die 
Vortrage,  die  anipiich  des  Internationalen  Workshops  "Subal- 
plne  Stone  Pines  and  Their  Environment — The  Status  of  Our 
Knowledge"  gehalten  wurden.  Dieser  Workshop  fand  vom 
5.  bis  1 1 .  Septennber  1 992  in  St.  Moritz  (Schweiz)  statt.  Die 
funf  subalpinen  Steinkieferarten  sind  ausschliepiich  in  der 
Nordhemisphare  verbreitet — davon  eine  in  Nordamerika 
und  vier  in  Eurasien.  Diese  Arten  weisen  viele  gemeinsame 
Merkmale  auf  und  besetzen  auch  uhnliche  okologische 
Nischen.  Unter  Steinkiefern  versteht  man  die  Arten  des 
Genus  Pinus  mit  funfnadeligen  Kurztrieben,  flugellosen 
Samen  und  Zapfen,  die  im  Reifezustand  geschlossen 
bleiben. 

Die  Vorgeschichte  dieses  Internationalen  Workshops 
beginnt  1 987.  Damals  fand  in  Bozeman  ein  Workshop  uber 
die  Whitebark  Pine  {Pinus  albicaulis)  statt,  dem  1989  ein 
gropes  Symposium,  ebenfalls  in  Bozeman,  folgte.  Beide 
Veranstaltungen  stiepen  auf  eine  grope  Resonanz  und  fuhrten 
zu  einem  wachsenden  Interesse  an  den  Hochgebirgsoko- 
systemen  des  westlichen  Nordamerika.  Eine  Vortragssektion 
dieses  Symposiums  war  den  Hochgebirgsokosystemen  der 
Welt  gewidmet.  Dabei  zeigte  sich,  dap  aus  einem  Vergleich 
mit  Steinkiefer-Okosystemen  anderer  in  ihrer  Naturaus- 
stattung  mit  dem  nordamerikanischen  Verbreitungsareal  der 
Whitebark  Pine  vergleichbarer  Regionen  der  Erde  auch 
wichtige  Informationen  zum  besseren  Verstandnis  der 
Whitebark  Pine  -  Walder  gewonnen  werden  konnen.  Als  ein 
wesentlicher  Unterschied  zu  den  amerikanischen  Whitebark 
Pine  -  Waldern  erwies  sich  die  zum  Tell  schon  einige  tausend 
Jahren  zuruckreichende  anthropogene  Beeinflussung  der 
eurasiatischen  Steinkiefernwalder  durch  den  Menschen.  Mit 
Sicherheit  lassen  sich  aus  den  dortigen  Verhaltnissen  auch 
nutzliche  Hinweise  auch  potentielle  Auswirkungen  men- 
schlicher  Eingriffe  in  den  amerikanischen  Whitebark  Pine  - 
Waldern  gewinnen. 

Anfang  1990  traf  sich  dann  eine  kleine  Gruppe  von  Wissen- 
schaftlern  und  Forstleuten  aus  den  westlichen  Vereinigten 
Staaten  mit  Dr.  Holtmeier  (Munster,  Deutschland)  in  Missoula 
(Montana),  um  uber  weiterfuhrende  Forschungen  iiber  die 
subalpinen  Steinkiefern  zu  beraten.  Der  nachste  Schritt 
sollte  eine  Verbesserung  und  Intensivierung  der  wissen- 
schaftlichen  Zusammenarbeit  der  Fachleute  aus  Europa, 
Asien  und  Nordamerika  sein.  Man  kam  uberein,  ein  en 
Internationalen  Workshop  zu  veranstalten,  mit  dem  Ziel,  dort 
Fragen  von  gemeinsamem  Interesse  zu  diskutieren  und  neue 
Forschungsansatze  zu  entwickeln.  Unmittelbar  nach  diesem 
Treffen  begannen  die  Vorbereitungen  zu  diesem  Workshop, 
der  dann  zweieinhalb  Jahre  spater  in  St.  Moritz  (Schweiz) 
stattfand. 

Seine  Hauptziele  waren: 

1 .  Verbesserung  und  Intensivierung  der  internationalen 
Zusammenarbeit. 

2.  Austausch  von  Forschungsergebnissen. 

3.  Forschungsbedarf  hinsichtlich  der  Okologie  und 
Bewirtschaftung  subalpiner  Steinkiefernwalder  aufzuzeigen. 

Uber  alle  funf  Steinkieferarten  wurden  Vortrage  gehalten. 
Insgesamt  waren  es  48  Referate,  an  denen  Wissenschaftler 
aus  12  Staaten  beteiligt  waren. 


Art 

Arve  (Synonym:  Zirbe) 
{Pinus  cembra) 

Sibirische  Arve 
{Pinus  sibirica) 


Sibirische  Zwergarve 
(Japanische 
Steinkiefer) 
{Pinus  pumila) 

Koreanische  Steinkiefer 
{Pinus  lioraiensis) 

Weiprinden-Kiefer 
{Pinus  albicaulis) 


Naturliches 
Verbreitunggebiet 

Europaische  Alpen,  Karpaten 


Russisch  Sibirien  und  ndrdliche 
Mongolei  (95  percent  in 
Rupiand;  5  percent  in  der 
Mongolei) 

Japan,  Korea,  Ostsibirien 


Nordost-China,  Nordkorea, 
Honshu,  Siidostasien 

Westliches  Nordamerika 


Folgende  Lander  waren  vertreten  (Zahl  der  Beitrage  in 
Klammern):  Rupiand  (9),  China  (1),  Japan  (4),  Schweiz  (5), 
Osterreich  (4),  Italien  (3),  Frankreich  (1),  Yugoslawien  (1), 
Tschechoslowakei  (1),  Rumanien  (1),  Deutschland  (3)  und 
USA  (15) 

Die  Hauptthemen  waren: 

•  Evolution  und  Taxonomie. 

•  Autokologie  der  subalpinen  Steinkieferarten. 

•  Wachstumscharakteristik. 

•  Einflup  der  Standortfaktoren. 

•  Regeneration. 

•  Bedeutung  der  Steinkiefern  fiir  die  Tiere. 

•  Waldstrukturen  und  -dynamik. 

•  Waldbewirtschaftung  und  -pflege. 

•  Forschungsbedarf. 

Die  Steinkiefern  stellen  einen  hohen  Anteil  an  der  Wald- 
bedeckung  groper  Gebiete.  Allein  Pinus  sibirica  bedeckt  ein 
Areal  von  40  Millionen  Hektar.  Der  gropte  Teil  davon  liegt  in 
Sibirien,  wo  der  Sibirischen  Arve  sowohl  in  ihrer  Bedeutung 
fur  die  Wirtschaft  als  auch  im  Hinblick  auf  die  biologische 
Situation  die  Rolle  eines  Schlusselfaktors  zukommt.  Pinus 
koraiensis  stellt  ein  Viertel  der  Holzproduktion  in  China. 
Pinus  albicaulis  ist  unter  anderem  fur  den  Grizzly-Bar  von 
existentieller  Bedeutung.  Auperdem  erfullt  diese  Art  in  hoch- 
gelegenen  Wassereinzugsgebieten  eine  wichtige  Schutz- 
funktion  und  erhoht  zudem  wesentlich  den  asthetischen  Reiz 
der  Gebirgslandschaft.  Pinus  cembra  kommt  in  Europa  eine 
wichtige  Rolle  bei  der  Lawinenvorbeugung  durch  Schutz- 
walder  zu.  Zugleich  liefert  sie  ein  fiir  Tafelungen  und  Schnitz- 
erein  geschatztes  Holz.  Auch  sie  ist  von  hohem  asthetischen 
Wert.  Pinus  pumila  tragt  in  Japan  und  Kamtschatka  ent- 
scheidend  zur  Stabilisierung  steiler  Gebirgshange  bei.  Die 
vielen  Gemeinsamkeiten  der  genannten  Arten  haben  ihre 
Ursache  vermutlich  in  ihrer  Abstammung  von  einer  Art, 
moglichenweise  von  Pinus  sibirica.  Charakteristisch  ist  fiir  sie 
auch  die  Verbreitung  ihrer  Samen  durch  den  nordamerika- 
nischen und  europaischen  Tannenhaher.  Zwischen  den 
subalpinen  Steinkiefern  und  den  Hahern  hat  sich  im  Laufe 
der  Evolution  ein  echter  Mutualismus  entwickelt. 


Der  Workshop  wurde  unterstutzt  durch  die  Intermountain 
Research  Station,  USDA  Forest  Service,  USA,  die  West- 
falische  Wilhelms-Universitat  (MOnster,  Deutschland),  die 
Eidgenossische  Forschungsanstait  fur  Wald,  Schnee  und 
Landschaft  (Schweiz)  sowie  durch  die  Karl  und  Sophie  Bind- 
ing -  Stiftung. 

Wir  danken  alien  Personen  und  Organisationen,  die  mit 
ihrem  Enthusiasmus  und  ihrer  tatkraftigen  Unterutzung  die 
Planung  und  die  Durchfuhrung  dieses  Workshops  ermbglicht 
haben. 

Wahrend  des  Workshops  fanden  unter  der  Leitung  der 
Gastgeber  drei  Exkursionen  statt.  Die  erste  fuhrte  in  die 
Larchen-Arvenwalder  oberhalb  des  bekannten  Engadiner 
Hohenkurortes  St.  Moritz.  Hauptthemen  waren  die  Ver- 
jungung  und  Sukzession  in  Lawinenzugen  und  im  Wald- 
grenzbereich,  die  Einbindung  von  Eichhornchen  und  Tannen- 
haher  in  das  Okosystem  des  Larchen-Arvenwaldes,  die 
Einflusse  von  Beweidung  und  Wild  sowie  die  Belastungen 
der  Walder  und  der  Landschaft  durch  den  modernen 
Tourismus. 

Die  zweite  Exkursion,  die  sich  im  wesentlichen  mit  okologi- 
schen  Problemen  und  Techniken  bei  der  Hochlagenauf- 
forstung  und  der  Behandlung  von  Lawineneinzugsgebieten 
befapte,  fuhrte  in  die  im  Waldgrenzbereich  gelegenen  Ver- 
suchsflachen  am  Stillberg  im  Dischmatal  bei  Davos.  Dort 
wurden  kunstliche  Lawinenverbauungen  und  Wieder- 
aufforstungen  mit  Arven,  Larche  und  Legfohre  als  Mittel  zur 
Lawinenvorbeugung  demonstriert. 

Die  dritte  Tour,  die  unter  dem  Thema  "subalpine  und  alpine 
Lebensraume"  stand,  fuhrte  zunachst  in  das  Morteratsch-Tal. 
Dort  ging  es  vor  allem  um  die  Vegetationssukzession  im 
Vorfeld  des  allmahlich  zuruckweichenden  Morteratsch- 
Gletschers  unter  besonderer  Berucksichtigung  der 
Besiedlung  durch  Arve  und  Larche.  Mit  einem  Besuch  der 
Diavolezza  am  Berninapap  fand  die  Exkursion  ihren 
Abschlup.  Von  der  Bergstation  (2973  m)  der  Luftseilbahn  bot 
sich  ein  uberwaltigender  Blick  auf  die  stark  vergletscherte 
Nordseite  der  Berninagruppe. 

Fur  alles,  die  eriebnisreichen  und  lehrrreichen  Exkur- 
sionen sowie  die  Veranstaltungen  in  St.  Moritz,  schulden  wir 
den  Gastgebern  —  Drs.  W.  Schonenberger,  R.  Hasler, 
J.  Senn,  R.  Lassig,  F.-K.  Holtmeier  und  Herr  U.  Wasem  — 
sowie  den  Engadiner  Fbrstern  J.  Altmann,  I. A.  Bisaz,  und 
G.  Bott  besonderen  Dank.  Nicht  zuletzt  geht  unser  Dank 
auch  an  Herrn  Prof.  Dr.  H.  Steinlin,  der  als  Vertreter  der 
Binding-Stiftung  selbst  an  dem  Workshop  teilnahm.  Sie  alle 
haben  mit  ihrem  gro^en  Engagement  bei  der  Planung  und 
der  Durchfuhrung  zum  Gelingen  dieses  ersten  internationalen 
Workshops  uber  die  Steinkiefern  beigetragen. 

Besonderer  Dank  gilt  auch  Frau  Kathy  McDonald  und 
Herrn  Richard  Klade  (Intermountain  Research  Station,  USA) 
fur  die  redaktionelle  Bearbeitung  der  Manuskripte. 

Das  Organistaionskomitee  setzte  sich  zusammen  aus: 

Dr.  W.  Schonenberger 
Dr.  W.  Hasler 
Dr.  J.  Senn 
Eidgenossische  Forschungsanstait 
fur  Wald,  Schnee  und  Landschaft 
Zurcher  Strape  1 1 1 
Ch-8093  Birmensdorf 
Schweiz 


Prof.  Dr.  F.-K.  Holtmeier 
Westfalische  Wilhelms-Universitat 
Institut  fur  Geographie 
Abteilung  Landschaftsokologie 
Robert-Koch-Strape  29 
D-48149  Munster 
Deutschland 

Dr.  Wyman  C.  Schmidt 
USDA  Forest  Service 
Intermountain  Research  Station 
Forestry  Sciences  Laboratory 
Montana  State  University 
Bozeman,  Montana,  USA  59717-0278 

Die  Vorbereitung  der  Manuskripte  fur  die  redaktionelle 
Bearbeitung  wurde  von  Dr.  Holtmeier  und  Dr.  Schmidt 
durchgefuhrt. 

Wahrend  des  Workshops  veranschiedeten  die  Teilnehmer 
eine  Resolution,  die  eine  noch  weitergehende  Koordination 
der  Forschung  und  bessere  gegenseitige  Information  uber 
neue  Forschungsergebnisse  zum  Ziel  hat.  Sie  umfapt 
folgende  Punkte: 

1 .  Grundung  einer  Internationalen  Vereinigung  fur  die 
Erforschung  der  Steinkiefern. 

2.  Bildung  eines  Internationalen  Organisationskomitees, 
das  erstmals  1993  (oder  spater)  in  Tomsk  (Rupiand) 
zusammentreten  soli,  um  die  kunftigen  Ziele  dieser  Ver- 
einigung zu  formulieren  und  eine  Organisationstruktur  zu 
schaffen. 

3.  Planung  des  nachsten  Internationalen  Workshops  uber 
die  Steinkiefern. 

4.  Durchfuhrung  dieses  Workshops  voraussichtlich  1995 
(Oder  spater)  in  Tomsk. 

Die  Welt  sieht  sich  vielen  Herausforderungen  durch 
Umweltprobleme  und  Ressourcennutzung  gegenuber. 
Davon  sind  auch  die  Gebirgswalder  und  borealen  Walder  der 
Nordhemisphare  betroffen,  zu  denen  auch  die  Steinkiefern- 
walder  gehoren.  Das  Wissen  uber  sie  ist  noch  luckenhaft. 
Indem  wir  unser  Wissen  zusammentragen  bilden  wir  eine 
solide  Grundlage,  von  der  aus  es  leichter  sein  wird,  tiefere 
Einsichten  in  die  okologischen  Wirkungsgefuge  dieser 
Waldokosysteme  zu  gewinnen.  Der  Internationale  Workshop 
in  St.  Moritz  mit  den  hier  zusammengestellten  Beitragen  ist 
ein  erster  Schritt  in  diese  Richtung.  Jetzt  geht  es  darum, 
diesen  Weg  weiterzugehen. 


Wyman  C.  Schmidt  F.-Kari  Holtmeier 


CONTENTS 


AUTHORS  TITLE  PAGE 

Wyman  C.  Schmidt  Distribution  of  Stone  Pines  1 

FIELD  TOURS 

Swiss  Hosts  Tour  Descriptions  8 

Mihailo  Grbic 

Friedrich-Karl  Holtmeier  Introduction  to  the  Upper  Engadine  and  Its  Forest  9 

EVOLUTION  AND  TAXONOMY 
Konstantin  V.  Krutovskli  Genetic  Differentiation  and  Phylogeny  of  Stone  Pine  Species  Based  on 

Dmitri  V.  Politov  Isozyme  Loci  19 

Yuri  P.  Altukhov 

Hermann  Mattes  Coevolutional  Aspects  of  Stone  Pines  and  Nutcrackers  31 

Dmitri  V.  Politov  Allozyme  Polymorphism,  Heterozygosity,  and  Mating  System  of  Stone  Pines  36 

Konstantin  V.  Krutovskii 

Diana  F.  Tomback  Genetic  Population  Structure  and  Growth  Form  Distribution  in  Bird-Dispersed  Pines  43 

William  S.  F.  Schuster 

BASIC  ECOLOGY 

Ernst  Frehner  Experiences  With  Reproduction  of  Cembra  Pine  52 

Walter  Schonenberger 

P.  L.  Gorchakovsky  Distribution  and  Ecology  of  Siberian  Stone  Pine  in  the  Urals  56 

Rudolf  Hasler  Ecophysiological  Investigations  on  Cembran  Pine  at  Timberline  in  the  Alps,  An  Overview  61 

Peter  A.  Khomentovsky  A  Pattern  of  Pinus  pumila  Seed  Production  Ecology  in  the  Mountains  of  Central 

Kamtchatka  67 

Takayoshi  Koike  Needle  Longevity  and  Photosynthetic  Performance  in  Cembran  Pine  and  Norway  Spruce 

Rudolf  Hasler  Growing  on  the  North-  and  East-Facing  Slopes  at  the  Timberline  of  Stillberg  in  the 

Hans  Item  Swiss  Alps  78 

Luo  Ju  Chun  The  Broad-leaved  Korean  Pine  Forest  in  China  81 

Tad  Weaver  Climates  Where  Stone  Pines  Grow,  A  Comparison  85 

GROWTH  CHARACTERISTICS 

Kurt  Holzer  Growth  of  Swiss  Stone  Pines  That  Originated  From  and  Were  Planted  at  Several 

Altitudes  in  the  Austrian  Alps  91 

Takuya  Kajimoto  Seasonal  Patterns  of  Growth  and  Photosynthetic  Activity  of  Pinus  pumila  Growing  on  the 

Kiso  Mountain  Range,  Central  Japan  93 

Herbert  Kronfuss  Height  Growth  in  Cembran  Pine  as  a  Factor  of  Air  Temperature  99 

Josef  Senn  Survival  and  Growth  of  Planted  Cembran  Pines  at  the  Alpine  Timberline  105 

Walter  Schbnenberg 
Ueli  Wasem 


AUTHORS  TITLE  PAGE 

INFLUENCE  OF  ENVIRONMENTAL  FACTORS 

Mihailo  Grbic  Changes  of  Swiss  Stone  Pine  Aphid  Life  Cycle,  Density,  and  Population  Structure  in 

High-Altitude  Swiss  Stone  Pine  Afforestation   111 

Raymond  J.  Hoff  Genetic  Consequences  and  Research  Challenges  of  Blister  Rust  in  Whitebark  Pine 

Susan  K.  Hagle  Forests  118 

Richard  G.  Krebill 

Karel  Kahik  Performance  of  Pinus  cembra,  P.  peace,  and  P.  strobiformis  Within  Air-Polluted  Areas   127 

Todd  Kipfer  Competition  and  Crown  Characteristics  of  Whitebark  Pine  Following  Logging  in 

Katherine  Hansen  Montana,  U.S.A   130 

Ward  McCaughey 

Penelope  Morgan  Fire  Ecology  of  Whitebark  Pine  Forests  of  the  Northern  Rocky  Mountains,  U.S.A   136 

Stephen  C.  Bunting 
Robert  E.  Keane 
Stephen  F.  Arno 

Tad  Weaver  Vegetation  Distribution  and  Production  in  Rocky  Mountain  Climates — With  Emphasis 

on  Whitebark  Pine  142 

REGENERATION 

I.  Blada  Variation  in  Size  and  Weight  of  Cones  and  Seeds  in  Four  Natural  Populations  of 

N.  Popescu  Carpathian  Stone  Pine  154 

Mitsuhiro  Hayashida  Role  of  Nutcrackers  on  Seed  Dispersal  and  Establishment  of  Pinus  pumila  and 

P.  pentaphylla  1 59 

Harry  E.  Hutchins  Role  of  Various  Animals  in  Dispersal  and  Establishment  of  Whitebark  Pine  in  the 

Rocky  Mountains,  U.S.A  163 

Hermann  Mattes  Size  of  Pine  Areas  in  Relation  to  Seed  Dispersal  172 

Ward  W.  McCaughey  The  Regeneration  Process  of  Whitebark  Pine   179 

Ikuko  Nakashinden  Japanese  Stone  Pine  Cone  Production  Estimated  From  Cone  Scars,  Mount 

Kisokomagatake,  Central  Japanese  Alps   188 

Diana  F.  Tomback  Effects  of  Seed  Dispersal  by  Clark's  Nutcracker  on  Early  Postfire  Regeneration  of 

Whitebark  Pine   193 

Vladislav  N.  Vorobjev  New  Trend  in  Dendrochronology:  1.  Theoretical  Principles  of  Reprochronology   199 

V.  N.  Vorobjev  New  Trend  in  Dendrochronology:  2.  Method  of  Retrospective  Study  of  Seminiference 

S.  N.  Goroshkevich  Dynamics  in  Pinaceae  201 

D.  A.  Savchuk 

IMPORTANCE  TO  WILDLIFE 

Ronald  M.  Lanner  Nutritive  Value  of  Whitebark  Pine  Seeds,  and  the  Question  of  Their  Variable  Dormancy  206 

Barrie  K.  Gilbert 

David  J.  Mattson  Bear  Use  of  Whitebark  Pine  Seeds  in  North  America  212 

Daniel  P.  Reinhart 

Diana  F.  Tomback  Ecological  Relationship  Between  Clark's  Nutcracker  and  Four  Wingless-Seed  Strobus 

Pines  of  Western  North  America  221 


AUTHORS  TITLE  PAGE 

FOREST  STRUCTURE  AND  DYNAMICS 

Giovanna  De  Mas  Structure  of  Swiss  Stone  Pine  Stands  in  Northeastern  Italy  226 

Elena  Piutti 

S.  Yu.  Grishin  Role  of  Pinus  pumila  in  Primary  Succession  on  the  Lava  Flows  of  Volcanoes  of 

Kamchatka  240 

Robert  E.  Keane  Decline  of  Whitebark  Pine  in  the  Bob  Marshall  Wilderness  Complex  of  Montana,  U.S.A  245 

Penelope  Morgan 

Renzo  Motta  Some  Aspects  of  Cembran  Pine  Regeneration  in  the  Italian  Cottian  Alps  254 

Alberto  Dotta 

Pietro  Piussi  Mixed  Cembran  Pine  Stands  on  the  Southern  Slope  of  the  Eastern  Alps  261 

Siegfried  Sauermoser  Current  Distribution  of  Cembra  Pine  in  the  Lechtal  Alps  269 

E.  P.  Smolonogov  Geographical  Differentiation  and  Dynamics  of  Siberian  Stone  Pine  Forests  in  Eurasia  275 

Use  WI6rick  The  Cembran  Pine  in  the  French  Alps:  Stand  Dynamics  of  a  Cembran  Pine  Forest 

in  Tueda  (Savoy,  France)  280 

FOREST  MANAGEMENT 

Werner  Frey  Silvicultural  Treatment  and  Avalanche  Protection  of  Swiss  Stone  Pine  Forests  290 

Riet  Gordon  Importance  and  Silvicultural  Treatment  of  Stone  Pine  in  the  Upper  Engadine  (Grisons)  294 

Jorg  Heumader  Cultivation  of  Cembran  Pine  Plants  for  High-Elevation  Afforestations  298 

Katherine  C.  Kendall  Whitebark  Pine  Conservation  in  North  American  National  Parks  302 

RESEARCH  NEEDS 

Wyman  C.  Schmidt  Research  Needs  in  Whitebark  Pine  Ecosystems  309 

Vladislav  N.Vorobjev  Problems  of  Comprehensive  Investigation,  Utilization,  and  Reproduction  of 

Nina  A.  Vorobjeva  Russian  Cedar  Pine  Forests  314 


DISTRIBUTION  OF  STONE  PINES 

Wyman  C.  Schmidt 


The  five  stone  pine  species  that  are  the  subject  of  this 
proceedings  occur  only  in  the  Northern  Hemisphere — one 
species  in  North  America  and  the  other  four  in  Europe 
and  Asia.  Taxonomically  they  fall  in  section  Strobus  and 
subsection  Cembrae.  They  are  characterized  by  having 
five-needled  fascicles,  cones  that  do  not  open  at  maturity, 
and  seeds  that  are  wingless. 

The  distributions  of  the  five  species  are  reasonably  well 
defined,  but  updates  on  their  distribution  usually  show 
minor  changes.  Also,  in  some  cases,  such  as  with  white- 
bark  pine  (Pinus  albicaulis),  the  species  occurs  only  at 
high-elevation  sites,  and  there  may  be  long  distances  be- 
tween the  mountain  ranges  that  support  this  species.  This 
species  ocoirs  over  a  large  geographic  area;  however,  there 
are  usually  large  intervening  valleys  and  lowlands  not 
suitable  for  whitebark  pine  forests.  The  same  is  true  for 
the  other  stone  pines.  Therefore,  just  because  a  distribu- 
tion boundary  of  a  species  encompasses  a  large  area,  it 
does  not  mean  that  all  the  mountains  and  valleys  are 
fully  occupied  by  that  species. 

For  the  purposes  of  this  proceedings,  we  have  chosen  to 
show  the  general  distribution  on  the  maps  of  the  five  spe- 
cies and  do  not  necessarily  include  all  of  the  individual 
outliers. 

The  distribution  maps  that  follow  are  adapted  fi'om  those 
pubUshed  by  Critchfield  and  Little  (1966).  The  maps  show 
that  the  five  species  are  mostly  discrete  populations,  but 
there  are  overlapping  distributions  of  Siberian  stone  pine 
(Pinus  sibirica)  and  Japanese  stone  pine  (P.  pumila),  and 
an  overlap  of  Japanese  stone  pine  and  Korean  stone  pine 
(P.  koraiensis). 

Whitebark  pine  occurs  only  in  North  America  at  high 
elevations  fi*om  central  British  Columbia  south  through 
the  Cascade  Mountains  of  Washington  and  Oregon  to  the 
southern  Sierra  Nevada  of  California,  in  a  limited  area  of 
V   Nevada,  and  extensively  in  Montana,  Idaho,  and  western 
Wyoming.  It  occurs  just  below  the  timberline  and  extends 
downward  elevationally  into  associations  with  several 
other  conifers. 

Swiss  stone  pine  grows  at  high  elevations  in  the  Alps  and 
Carpathian  Moimtains.  It  occurs  mostly  in  Switzerland, 
Austria,  northern  Italy,  southeastern  France,  Romania, 
Czechoslovakia,  and  Poland.  Although  Swiss  stone  pine 
is  considered  closely  related  to  Siberian  stone  pine,  over 
2,000  kilometers  presently  separate  the  two  species. 


Swiss  stone  pine  commonly  grows  in  association  with 
other  conifers. 

Siberian  stone  pine  is  widely  distributed  in  Eurasia  fi"om 
the  Ural  Moimtains  through  western  and  central  Siberia  to 
northern  MongoUa — an  east-west  distance  of  about  4,000 
kilometers.  It  overlaps  the  range  of  Japanese  stone  pine 
on  the  eastern  portion  of  its  distribution  near  Lake  Baikal 
in  Siberia.  Different  than  both  whitebark  pine  and  Swiss 
stone  pine,  which  only  occupy  high-elevation  sites,  Sibe- 
rian stone  pine  (usually  called  cedar-pine  or  Siberian  ce- 
dar by  Russians)  commonly  grows  in  the  plains  and  river 
valleys,  as  well  as  in  the  mountain  areas. 

Japanese  stone  pine  grows  in  northeastern  Siberia  and 
in  subalpine  parts  of  Japan,  Korea,  and  Manchvuia.  It  ex- 
tends north  almost  to  the  Arctic  Ocean  (70°  N.),  west  to 
Mongolia  and  to  the  Lake  Baikal  area,  and  south  to  Korea 
and  Honshu,  Japan.  It  grows  from  low  elevations  in  the 
north  to  high  elevations  in  the  south.  It  commonly  forms 
thickets  just  below  barren  tundra  on  tops  of  mountains  or 
lowlands  in  the  extreme  north.  Like  Siberian  stone  pine, 
it  extends  nearly  4,000  kilometers  fi*om  east  to  west  and 
3,000  kilometers  north  to  south. 

Korean  stone  pine  occurs  throughout  Korea  and  eastern 
Manchuria  into  southeastern  Siberia  as  well  as  on  the  is- 
lands of  Honshu  and  Shikoku  in  Japan.  It  tends  to  grow 
in  more  maritime  conditions  and  lower  moimtains  than  its 
neighbor — the  Japanese  stone  pine.  Its  geographic  range 
overlaps  that  of  the  Japanese  stone  pine,  but  the  species 
occupy  different  ecological  niches.  Korean  stone  pine,  im- 
like  its  counterparts,  will  grow  in  association  with  hard- 
woods where  near-maritime  cUmates  prevail.  Its  valuable 
wood  makes  it  a  much  sought  after  species. 

The  following  maps  illustrate  the  geographic  distribu- 
tion of  the  five  stone  pines.  These  descriptions  are  based 
primarily  on  those  of  Critchfield  and  Little  (1966)  and 
Mirov  (1967). 

References 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agricvdture,  For- 
est Service.  97  p. 

Mirov,  N.  T.  1967.  The  genus  Pinus.  New  York:  Ronald 
Press  Company.  602  p. 


1 


NORTH  AMERICA 


135°  130°  125°  120°  115°  110° 


130°  125°  120°  115°  110°  105°  100° 


2 


EUROPE 


10°  5°  0°  5°  10°        15°  20°  25°  30°  35° 


0°  5°  10°  15°  20°  25°  30°  35° 


3 


ASIA 


4 


ASIA 


5 


ASIA 


Field  Tours 


International  Workshop 
St.  Moritz  1 992 


Field  tours  were  an  important  component  of  the  Stone  Pine  Workshop.  First, 
the  tours  provided  an  informal  setting  for  scientists  from  the  12  countries  to  become 
acquainted  with  each  other  and  their  research;  second,  the  tours  demonstrated  the 
research  and  management  of  Swiss  stone  pine  forests  in  the  rugged  and  spectacularly 
beautiful  mountains  of  Switzerland. 

The  following  section  is  the  handout  from  the  Swiss  hosts  used  to  describe 
the  three  major  field  tours.  The  drawings  by  a  gifted  scientist,  Mihailo  Grbic,  from 
Belgrade,  Yugoslavia,  depict  a  symbolic  participant,  Dr.  Stone  Pine,  enjoying  the  rather 
strenuous  excursions.  Grbic  also  created  the  welcome  drawing  used  as  sections 
division  indicators  in  this  publication. 


7 


TOUR  DESCRIPTIONS 

Swiss  Hosts 


We  will  try  to  use  pleasant  weather  for  the  field  tours. 
Therefore  this  preliminary  schedule  may  change  a  little. 
Usually  weather  is  fine  in  September.  However,  snow  and 
fi-ost  are  possible.  Do  not  forget  warm  clothing,  tough  foot- 
wear, backpack,  sunblock,  and  simglasses.  During  the  field 
toiu-s  we  will  have  lunch  fi-om  the  backpack. 

Engadine  Cembra  Pine-Larch  Forest 

Sunday  6th.  St.  Moritz,  Muottas  da  Schlarigna.  Walk 
fi-om  St.  Moritz  Bad  (Hotel  "Laudinella'')-Stazer  Wald- 
Muottas  da  Schlarigna-St.  Moritz  Bad  (1,800  to  2,300  m 
above  sea  level  [a.s.l.]). 

Cembra  pine-larch  forest:  composition,  structure,  growth 
characteristics,  old  cembra  pines,  regeneration  pattern, 
understory  vegetation,  avalanche  tracks  with  green  alder 
{Alnus  viridis)  formation,  influence  of  seed-caching  ani- 
mals (squirrel,  nutcracker),  effects  of  seed  dispersal  by 
the  thick-billed  European  nutcracker  (Nucifraga  c. 
caryocatactes). 

Human  influence  on  the  forest  distribution  pattern, 
human-caused  timberline,  influence  of  grazing  livestock 
and  game,  and  of  toiu"ism;  invasion  of  rarely  used  and 
abandoned  alpine  and  forest  pastures  by  trees,  influence 
of  microtopography  and  microclimate  on  alpine  vegetation 
and  on  growth  of  invading  trees. 

Ecology  and  Technique  of  High- 
Altitude  Afforestation 

Tuesday  8th.  Davos,  Stillberg.  Bus  fi"om  St.  Moritz- 
Albula  pass-Davos-Teufi  (Dischma  Valley,  1,700  m), 
5-hour  walk  to  and  through  Stillberg  experimental  area 
near  timberline  at  2,000  to  2,230  m  a.s.l.  and  back  to  the 
valley  bottom  (Teufi). 

The  experimental  area  lies  on  a  steep  avalanche  catch- 
ment slope.  Construction  of  avalanche  barriers.  Dam- 
ages, diseases,  mortality  and  development  of  a  17-year-old 


plantation  of  Pinus  cembra,  Larix  decidua,  and  P.  mugo. 
Influence  of  different  site  factors  such  as  wind,  tempera- 
ture, irradiation,  snow  movement,  duration  of  snow  cover, 
soil  properties,  vegetation  on  plantation  success.  Eco- 
physiology  of  subalpine  timberline  trees.  Conclusions  for 
plantation  technique  and  spatial  patterns. 


Subalpine  and  Alpine  Environments 

Friday  11th.  Morteratsch  Glacier,  Diavolezza. 
Morteratsch  Glacier  (2,000  m,  2-  to  3-hour  walk).  Train  to 
Morteratsch  station,  walk  to  Morteratsch  Glacier  via  Alp 
Chiinetta  (2,052  m);  old-growth  cembra  pine  forests,  re- 
generation, invasion  of  the  lateral  moraine  (fi-om  1850) 
and  the  glacier  forefield  by  cembra  pine  and  larch. 

Diavolezza  (2,973  m).  Train  fi-om  Morteratsch  station 
through  the  Bemina  Valley  to  Diavolezza  station  (2,200  m) 
on  Bemina  Pass,  and  by  ropeway  to  Diavolezza.  Magnifi- 
cent view  of  the  Bernina  Massive  (4,049  m)  with  Pers  Gla- 
cier and  Morteratsch  Glacier. 

★  ***★ 

In  addition  to  providing  valuable  scientific  and  manage- 
ment information,  the  introductory  paper  by  Friedrich- 
Karl  Holtmeier  that  follows  gave  background  and  baseline 
data  describing  the  local  Swiss  stone  pine  forests  that 
were  the  focus  of  the  field  tours. 

***** 


8 


INTRODUCTION  TO  THE  UPPER 
ENGADINE  AND  ITS  FOREST 

Friedrich-Karl  Holtmeier 


Abstract — The  Engadine  is  characterised  by  great  mean  eleva- 
tion (2,400  m)  and  a  relatively  continental  climate.  Cembra  pine 
{Pinus  cembra)  and  European  larch  (Larix  decidua)  are  the  most 
important  tree  species.  The  area  was  already  settled  in  prehis- 
toric time.  Since  then,  mountain  forest  has  been  influenced  by 
human  impact.  During  history  timberline  became  lower  by  about 
150-300  m.  At  present,  the  upper  forest  limit  is  located  at  about 
2,200-2,300  m.  Composition  and  structure  of  the  forest  also 
changed  in  that  larch  could  spread  at  the  cost  of  cembra  pine  on 
all  areas  that  were  suitable  for  grazing.  After  grazing  pressure 
has  declined  natural  succession  is  revived,  and,  in  the  long  term, 
larch  will  be  replaced  by  cembra  pine,  if  not  specially  managed. 
Natural  stands  of  Norway  spruce  (Picea  ahies)  occur  on  the 
northwest-exposed  slopes  of  the  main  valleys  within  an  area 
characterized  by  higher  humidity.  Spruce  reaches  its  upper  limit 
at  2,000  m.  Dwarfed  mountain  pine  (Pinus  mugo)  and  green  al- 
der (Alnus  viridis)  are  most  common  in  avalanche  tracks.  Green 
alder  is  to  be  found  mainly  on  moist  sites,  whereas  mountain 
pine  prevails  under  dry  conditions.  Alder  also  occurs  in  bogs. 


The  Engadine — Engiadina  in  Rhacto-Romsinic  lan- 
guages— is  located  in  the  eastern  part  of  Switzerland  and 
comprises  the  uppermost  drainage  area  of  the  Inn  River 
(fig.  1).  The  main  valley  trends  from  southwest  to  north- 
east. By  an  administrative  border  the  Engadine  is  divided 
into  Upper  and  Lower  Engadine.  However,  landscape 
physiognomy  of  both  sections  is  also  different.  The  Lower 
Engadine  is  deeply  cut  into  the  mountains,  and  mountain 
slopes  are  steep  and  nagged.  The  Upper  Engadine  main 
valley,  on  the  other  hand,  is  wide  and  topography  is  com- 
paratively smooth.  Between  St.  Moritz  and  the  Maloja 
Pass  the  valley  is  characterized  by  four  lakes  separated  by 
alluvial  plains.  To  the  southeast  the  Bemina  Valley  gives 
access  to  the  famous  glaciated  Bemina  Group. 

The  Engadine  is  characterized  by  great  mean  elevation 
(2,400  m).  The  bottom  of  the  main  valley  is  located  at  an 
altitude  of  about  1,800  m.  The  tributary  valleys  climb  up 
to  about  2,300  m.  The  highest  peaks'  elevations  are  about 
4,000  m  (Piz  Bernina  4,049  m;  fig.  1).  In  view  of  its  great 
elevation  the  Upper  Engadine  has  been  called  The  Roof 
of  Europe"  by  the  English. 

Due  to  the  geographical  location  in  the  central  Alps 
and  the  high  mass-elevation  of  the  Engadine  the  climate 
is  rather  continental,  characterized  by  a  relatively  great 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Friedrich-Karl  Holtmeier  is  Professor  of  Landscape  Ecology  and  Geogra- 
phy, Department  of  Geography,  University  of  Munster,  Germany. 


amplitude  of  mean  temperature  and  comparatively  low 
amount  of  precipitation  (fig.  2).  Annual  precipitation  de- 
creases from  about  1,400  mm  at  Maloja  Pass,  where  the 
main  valley  is  open  to  moisture-carrying  air  ciurents  from 
the  southwest  (fig.  3),  to  less  than  1,000  mm  at  St.  Moritz. 
At  timberline,  however,  precipitation  is  much  higher  than 
at  the  valley  bottom,  as  can  be  concluded  from  the  data 
available  for  Julier  Hospiz  on  Julier  Pass  and  for  Bemina 
Hospiz  on  Bemina  Pass.  At  both  stations,  anniial  precipi- 
tation exceeds  1,500  mm. 

Due  to  the  continental  climate,  the  upper  limits  of  veg- 
etation and  cultivation  reach  relatively  high  altitudes.  In 
the  Upper  Engadine,  forest  limit  is  located  at  about  2,200- 
2,300  m.  Solitary  crippled  trees  may  still  be  found  at  and 
even  above  2,500  m  (Holtmeier  1965,  1974). 

In  the  Upper  Engadine,  artifacts  were  found  that  give 
evidence  of  agriculttire  during  the  Bronze  Age  (about 
3,000  years  B.P.),  while  in  the  Lower  Engadine  hunters 
and  shepherds  were  already  present  in  the  Neohthic  Age 
(4,000  years  ago;  Conrad  1940).  Trading  and  mining  be- 
came increasingly  important  to  the  economy  dvuing  his- 
toric time.  Thus,  there  was  increasing  human  impact  on 
the  moimtain  ecosystems  throughout  history.  Almost  no 
untouched  nature  was  left,  although  the  kind  and  the  in- 
tensity of  the  himaan  disturbances  were  different. 

DEPRESSION  OF  TIMBERLINE 

Due  to  the  great  elevation  of  the  area  humans  could 
only  use  the  mountains  up  to  a  certain  altitude;  there  the 


Figure  1— Location  of  the  Upper  Engadine. 


9 


3000 


2500 


2000 


1500 


1000  - 


500  - 


mm 


Figure  2 — Annual  precipitation.  Note  the  de- 
crease from  Maloja  Pass  toward  St..  Moritz  and 
Schuls  (Lower  Engadine). 


Figure  4 — Zuoz  Village  (1 ,700  m).  The  original 
forest  was  almost  totally  removed  as  a  conse- 
quence of  alpine  pasturing  (upper  slope)  and 
agriculture  (artificial  terraces  above  the  village). 
Only  two  small  stands  were  left  on  the  steep 
north-exposed  slopes  of  the  narrow  tributary 
valleys  and  one  above  the  village  for  protection 
from  avalanches.  Photo  taken  1 968. 


main  influence  was  concentrated  in  the  forest  belt  not 
only  on  favorable  southern  exposures  but  also  on  all 
sections  that  were  fairly  accessible — trough  shoulders, 
terraces,  and  similar  almost-level  areas  even  in  north- 
exposed  locations.  During  former  times  cereals  were 
grown  up  to  elevations  of  about  2,300  m,  favorable  topog- 
raphy provided.  Many  of  the  high-altitude  forests  were 
cleared  for  pastoral  use,  while  those  at  lower  elevations 


gave  way  to  agriculture.  The  extent  of  timberline  depres- 
sion varied  in  relation  to  exposure  and  topography.  On 
some  easily  accessible  southern  exposures  of  the  main 
valley,  for  example,  and  of  the  Fex  Valley  and  Fedoz  Val- 
ley, the  forest  was  totally  removed  except  for  sections 
characterized  by  too  great  steepness  (figs.  4,  5).  Other  for- 
ests were  destroyed  by  ore  mining,  salt  works,  and  char- 
coal production,  especially  during  the  Middle  Ages.  On 


Figure  3 — View  from  Lej  dais  Chods  (northwest- 
facing  slope  of  the  main  valley)  southwest  to- 
ward the  gap  of  Maloja  Pass. 


Figure  5 — View  to  southeast-exposed  slope  of  the  main 
valley.  In  the  foreground,  Samaden  Village  (1 ,720  m). 
Forest  was  removed  from  all  gentle  topography  (trough 
shoulders,  lower  slope  area)  and  reduced  to  a  narrow 
belt  on  the  steep  trough  wall.  Photo  taken  1969. 


10 


Figure  6 — View  from  Diavolezza  trail  to 
Bernina  Pass  (2,328  m).  Bernina  Pass  was 
covered  by  larch-cembra  pine  forest  until  the 
Middle  Ages.  Then,  the  forest  became  a 
victim  of  mining.  Until  present,  the  area  is 
used  in  summer  for  grazing  cattle. 


the  Bernina,  Abula,  and  Jiilier  Passes  (about  2,300  m)  the 
forest  disappeared  as  a  consequence  of  mining.  These  ar- 
eas v^^ere  then  used  as  alpine  pastures  (fig.  6). 

The  upper  timberUne  receded  by  150-300  m  compared 
to  its  position  during  the  postglacial  thermal  optimvmi. 
That  timberline  depression  was  partly  due  to  a  general 
cooling.  However,  at  least  since  the  Middle  Ages  the  ef- 
fects of  the  deterioration  of  the  climate  on  subalpine  tree 
growth  have  been  compounded  or  even  superimposed  by 
the  human  influences  mentioned  earlier. 

COMPOSITION  AND  STRUCTURE 

As  a  consequence  of  these  activities  not  only  did  the  up- 
per timberline  become  lower  by  about  150-300  m,  but  also 
species  composition  and  structures  of  the  moimtain  for- 
ests changed  considerably  (Auer  1947;  Campell  1944; 
Holtmeier  1965,  1967).  Larch  (Larix  decidua)  and  cembra 
pine  {Pinus  cembra)  are  the  most  important  tree  species 
in  the  Upper  Engadine  forests.  These  forests  belong  to 
the  so-called  silicate  type  (Larici  Cembretum,  Ellenberg 
1978;  Larici-Pinetum  cembrae,  Oberdorfer  1970)  of  larch- 
cembra  pine  forest,  which  is  common  in  the  crystalline 
Central  Alps.  The  present  distribution  pattern  of  both 
species  in  the  Upper  Engadine  can  be  described  as  fol- 
lows: Cembra  pine  is  mainly  spread  on  steep  and  inacces- 
sible slopes  and  north  exposures,  whereas  larch  domi- 
nates on  easily  accessible  locations  exposed  to  the  south. 

The  influence  of  the  local  topography  on  the  distribu- 
tion pattern  is  most  pronovmced  in  the  trough-shaped 
Bever  Valley  and  Bernina  Valley  (fig.  7).  Pure  larch 
stands  are  confined  to  the  relatively  gentle  slope  areas 
(talus  cones,  screes,  boulder  fans,  etc.)  that  extend  below 
the  steep  and  rocky  trough  walls,  while  cembra  pine  pre- 
vails on  trough  walls  and  also  forms  the  upper  timberline. 
This  pattern  is  most  conspicuous  on  southern  exposures. 
On  north-exposed  slopes  cembra  pine  also  occurs  on  their 
lower  parts,  mixed  vnth  larch. 


This  distribution  of  larch  and  cembra  pine  is  mainly 
caused  by  human  influences,  especially  by  pastoral  use. 
Since  cembra  pine  impedes  the  growth  of  grass  and  herbs 
on  the  forest  floor,  while  larch  does  not,  cembra  pine  was 
cleared  on  slopes  suitable  for  grazing  and  restricted  to 
steep,  inaccessible  locations.  Moreover,  the  wood  of  cem- 
bra pine  was  used  for  many  purposes  such  as  paneling 
and  making  furniture,  milk  vessels,  or  carvings.  Cembra 
pine  also  is  more  susceptible  to  forest  fires  than  larch, 
which  is  protected  by  a  thick  corklike  bark.  As  indicated 
by  charcoal -rich  soil  horizons,  forest  fires  frequently 
occvured  in  the  past.  Probably,  many  of  them  were 
caused  by  shepherds  when  they  burned  alpine  pastures 
to  remove  "weeds"  such  as  Rhododendron  ferrugineum 
and  Juniperus  nana,  and  fires  ran  out  of  control.  Thus, 
for  many  reasons,  larch  co;ild  spread  at  the  cost  of  cembra 
pine,  as  was  also  the  case  in  many  other  central-alpine 
valleys.  At  many  places  the  high-elevation  forest  is  over- 
mature and  very  sensitive  to  environmental  impacts  such 
as  windthrow  and  snow  slides. 

CURRENT  CHANGES 

Due  to  modem  change  of  economic  structure — tourism 
has  become  the  main  base  of  the  economy  in  the 
Engadine — grazing  pressure  on  the  forests  declined,  and 
cembra  pine  is  invading  the  former  pasture  forest.  Larch, 
which  is  a  light-demanding  species,  is  going  to  be  out- 
competed  by  cembra  pine,  and  natural  succession  ft-om 
larch  to  cembra  pine  forests,  which  had  been  interrupted 
by  human  disturbances  for  some  hundred  years,  is  re- 
vived (Holtmeier  1967,  1990,  1993).  Larch  will  gradually 
be  replaced  by  cembra  pine,  if  sites  are  not  specially  man- 
aged by  exposing  mineral  soil  or  selective  cutting  of 
cembra  pine. 

Locally,  however,  invasion  of  cembra  pine  in  the  imder- 
story  is  hampered  by  mass  outbreaks  of  the  larch  bud 
moth  (Zeiraphera  diniana).  After  having  defoliated  the 
larch  crowns,  the  caterpillars  will  feed  on  the  cembra 
pines  in  the  understory  and,  together  with  secondary 
parasites,  cause  severe  damage  to  them.  Young  cembra 
pines  become  crippled  or  killed. 

In  forest  gaps  smaller  than  the  height  of  the  surround- 
ing trees,  accvunulation  of  snow  and  length  of  v^dnter  snow 
cover  may  critically  increase  (fig.  8),  and  consequently 
seedlings  and  saplings  of  cembra  pine  may  be  seriously 
damaged  by  snow  fungus  {Phacidium  infestans). 

Cembra  pine  also  is  invading  abandoned  or  only  rarely 
used  alpine  pastures.  Seeds  are  dispersed  by  the  nut- 
cracker {Nucifraga  caryocatactes),  occasionally  far  beyond 
the  present  forest  limit  (Holtmeier  1966,  1974;  Mattes 
1978,  1982).  Larch  seeds,  however,  are  prevented  from 
getting  into  a  suitable  seedbed  by  the  dense  alpine  vegeta- 
tion. Thus,  it  is  cembra  pine  that  is  the  pioneer  species 
on  former  pasture  areas  (fig.  9),  while  larch  only  occurs 
where  bare  mineral  soil  is  exposed.  Survival  of  the  invad- 
ing trees  mainly  depends  on  microclimates  and  resulting 
site  conditions.  They  considerably  deteriorated  when  the 
forest  was  removed  in  the  past.  Favorable  effects  of  the 
general  warming  on  tree  growth  are  superimposed  by 
unsuitable  local  site  conditions  (fig.  10).  Consequently, 
natural  reforestation  follows  the  favorable  sites  while 
other  sites  will  remain  treeless  probably  for  a  long  time. 


11 


Figure  7 — Distribution  of  tree  species  in  the  Bernina  Valley  (top)  and 
Bever  Valley  (bottom).  Both  maps  have  been  redrawn  from  Holtmeier 
(1967). 


12 


Figure  8 — Big  snow  masses  are  accumulated 
in  the  glade.  Incoming  solar  radiation  is  inter- 
cepted by  the  trees,  while  energy  loss  by  out- 
going long-wave  radiation  is  almost  unimpeded 
Due  to  these  conditions  snowmelt  is  consider- 
ably delayed,  and  young  cembra  pines  may  be 
infected  by  snow  fungus. 


As  to  successful  germination  and  survival  of  seedlings, 
seed  dispersal  by  the  nutcracker  is  much  more  effective 
than  wind-mediated  transport  because  many  seeds  are 
cached  at  sites  relatively  favorable  to  germination  and 
seedling  growth  (not  too  long  covered  with  snow).  On  the 
other  hand,  the  seed  caches  usually  provide  sufficient  soil 
moisture  because  they  are  2-4  cm  under  the  surface,  thus 
being  protected  from  direct  insolation.  Finally,  seeds  col- 
lected by  the  nutcracker  are  characterized  by  a  relatively 
high  germination  capacity  (Mattes  1978,  1982),  and  there 
is  almost  no  loss  of  cached  seeds  by  seed-eating  predators. 


Figure  9 — Cembra  pine  invading  an  abandoned 
alpine  pasture  (about  2,250  m)  on  the  northwest- 
exposed  slope  of  the  main  valley.  Invasion  fol- 
lows mainly  convex  topography,  while  sites 
covered  too  long  with  snow  and  characterized  by 
snow  creep  and  avalanches  remain  treeless. 


However,  due  to  high  density  of  seedling  clusters  some 
of  the  trees  will  become  victims  of  root  competition.  More- 
over, snow  accumulation  and  duration  of  the  snow  cover 
may  gradually  be  increased  by  the  influence  of  the  grow- 
ing trees  on  windflow  near  the  surface.  At  such  sites, 
yoimg  cembra  pines  covered  too  long  by  snow  may  be 
killed  by  snow  fungi  such  as  Phacidium  infestans 
(Holtmeier  1967,  1974,  1990  ). 

Natural  invasion  by  trees  (cembra  pine,  larch,  vallows, 
rowan,  and  also  spruce)  can  be  observed  in  front  of  and  on 
the  lateral  moraines  of  the  Morteratsch  Glacier  and  Roseg 
Glacier  (Holtmeier  1965,  1974,  1990;  fig.  11).  These  gla- 
ciers have  been  retreating  more  or  less  continuously  by 
about  30  m  per  year  since  the  beginning  of  our  century. 
Larch  alreadj  occurs  about  20  years  after  glacier  retreat, 
while  cembra  pine  ynW  follow  later.  Then,  however, 
cembra  pine  invades  these  areas  much  more  rapidly  than 
larch.  There  is  some  evidence  that  areas  which  became 
ice-free  about  50  to  70  years  ago  were  more  rapidly  in- 
vaded by  cembra  pine  than  those  from  where  the  glacier 
retreated  more  than  100  years  ago.  That  might  be  ex- 
plained by  mesoclimatic  conditions  improving  due  to  gla- 
cier retreat. 


7 


1.  Situation  during  the  postglacial  thermal  optimum 


forest  1  

 ecotone  1  alpine 

-  2100 

-  2100  ijpr 

forest  climate  replaced  by  the  very  locally 
changing  pattern  of  microclimates  ,  similar 
to  that  above  the  previous  tinfterline 

2.  Situation  at 

present 

Figure  10 — Change  of  microclimatic  conditions  in  the 
upper  Subalpine  after  the  uppermost  forest  was  re- 
moved by  humans. 


13 


1.5  km   1 


^Jk  cembrapine  ^^^i   larch  ''iV'    alpine  vegetation 

Figure  11 — Invasion  of  the  forefield  and  lateral  moraine  of  the  retreating  Morteratsch  Glacier  by  cembra 
pine  and  larch  (schematically). 


At  places,  however,  distribution  patterns  of  invading 
larches  and  cembra  pines  very  likely  depend  on  soil  mois- 
ture available,  as  can  be  observed  in  the  forefield  of 
Morteratsch  Glacier,  for  example  (fig.  11).  Obviously, 
relatively  dry  sites — such  as  the  well-drained  upper  part 
of  the  lateral  moraines  or  roche  moutonnee  areas — can  be 
invaded  by  cembra  pine  more  easily  than  by  larch,  be- 
cause seeds  of  cembra  pine  are  cached  by  the  nutcracker 
in  small  cracks  or  fine-soil  pockets  between  the  boulders 
where  sufficient  moisture  is  available  for  germination  and 
seedling  growth.  On  the  other  hand,  it  depends  totally  on 
chance  whether  wind-mediated  larch  seeds  will  get  into 
such  a  seedbed.  At  the  valley  bottom,  conditions  appear 
to  be  more  favorable  for  invasion  by  larch  because  of 
great  numbers  of  open  patches  of  bare  mineral  soil  cre- 
ated by  extensive  accumulation  of  fine  material  between 
the  boiilders.  At  such  sites  soil  moisture  is  usually  avail- 
able, and  there  is  a  better  chance  for  larch  seeds  to  reach 
a  suitable  seedbed. 


Spruce  (Picea  abies)  also  occurs  within  cembra  pine- 
larch  forest.  Natural  stands,  however,  are  confined  to  the 
northwest-exposed  slope  of  the  main  valley  between  the 
village  of  Sils  and  the  small  Lake  of  Staz  (fig.  12).  This 
area  is  characterized  by  increased  humidity  due  to  a  cloud 
belt  that  usually  appears  when  the  Upper  Engadine  is  in- 
fluenced by  moisture-carrying  air  currents  from  the 
southwest  (Holtmeier  1965,  1966,  1967).  The  upper  dis- 
tribution limit  of  spruce  is  located  at  about  2,000  m.  Al- 
though spruce  is  missing  between  Maloja  Pass  (1,815  m) 
and  Sils,  there  is  some  evidence  that  this  species  immi- 
grated from  the  south  across  Maloja  Pass  in  6000  to  5500 
B.C.  (Campell  1944;  Kleiber  1974).  The  gap  can  probably 
be  attributed  to  human  influences.  On  the  opposite  slope 
spruce  does  not  occur,  except  for  planted  trees  (Schlatter 
1935).  That  might  be  explained  by  lower  humidity  and 
lower  soil  moisture  on  that  slope  exposed  to  the  southeast. 
Anyway,  humans  also  could  have  removed  shade-giving 
spruce  (Holtmeier  1967).  After  deglaciation,  cembra  pine 
invaded  the  Upper  Engadine  earlier  than  spruce  (Kleiber 


Figure  12 — Distribution  of  spruce  {Picea  abies)  on  the  northwest-exposed  slope  of  the  main  valley  (from 
Holtmeier  1967,  modified). 


14 


Figure  13 — Arboreous  and  dwarfed  mountain  pine  in 
bog  near  Maloja  Pass. 


1974).  Otherwise  it  would  have  been  outcompeted  by  the 
latter  species — at  least  at  lower  elevations  (Heitz  1975; 
Mattes  1978). 

KRUMMHOLZ  SPECIES 

Dwarf  pine  {Pinus  mugo)  and  green  alder  (Alnus  viri- 
dis)  form  dense  and  locally  extended  krummholz  stands. 
The  krummholz  growth  of  both  these  species  is  genetically 
predetermined  and  not  climatically  induced.  That  makes 
them  different  from  climatically  trimmed,  flagged,  and 
matlike  "krummholz"  that  forms  the  forest-alpine  timdra 
ecotone  in  the  Rocky  Moimtains,  for  example  (Holtmeier 
1973,  1974,  1981). 

Moreover,  there  is  no  contiguous  Pinus  mugo  krumm- 
holz belt  above  the  high-stemmed  forest  as  is  peculiar  to 
the  northern  Alps.  In  the  Upper  Engadine,  both  Pinus 
mugo  and  Alnus  viridis  occur  mainly  on  slopes  endan- 
gered by  avalanches.  Their  elasticity  enables  them  to 
resist  avalanches  and  snow  slides  better  than  upright- 
growing  tree  species,  which  normally  will  be  eliminated 
at  such  sites.  Some  slopes,  where  the  original  forest  had 
been  destroyed  by  avalanches  and  fire,  were  completely 


afforested  with  moimtain  pine.  Moimtain  pine  is  also 
common  in  bogs  where  it  grows  on  slightly  convex  topog- 
raphy (mostly  rocky  outcrops).  At  such  sites  the  arbore- 
ous growth  form  also  occiu-s,  as  is  the  case  on  Maloja  Pass 
or  in  the  Lake  Staz  area,  for  example  (fig.  13). 

In  general,  Pinus  mugo  prevails  on  limestone  and  dolo- 
mite, while  Alnus  viridis  is  common  mainly  on  silicate 
material.  In  view  of  this  distribution,  both  species  are 
said  to  be  vicarious  plants.  In  the  Upper  Engadine,  how- 
ever, the  availability  of  soil  moistiu^e  seems  to  be  the  fac- 
tor controlling  the  distribution  pattern.  There,  Pinus 
mugo  is  more  common  on  southern  exposures,  whereas 
Alnus  viridis  is  mainly  spread  on  north-facing  slopes  pro- 
viding sufficient  soil  moistiire  and  fi'esh  soils  rich  in  nutri- 
ents. On  svmny  slopes,  however,  Alnus  viridis  is  confined 
to  moist  microsites,  as  alongside  rivlets,  for  example.  On 
northern  exposures  Pinus  mugo  successfully  competes 
with  Alnus  viridis  at  sites  relatively  dry  and  poor  in  nu- 
trients (fig.  14). 

Birch  {Betula  pubescens)  is  another  species  that  is  usu- 
ally confined  to  avalanche  tracks  and  bogs.  Birch  is  able 
to  recover  fi-om  breakage  by  thriving  basal  shoots.  There- 
fore, multistemmed  growth  is  quite  common  to  this  spe- 
cies (fig.  15). 

At  present,  the  Upper  Engadine  has  a  higher  forest 
cover  compared  to  150  years  ago  and  earlier.  Since  then 
many  afforestations  were  carried  out  to  close  the  gaps 
caused  by  historical  hiunan  impact.  Also  "exotic"  species 
were  used — such  as  Picea  pungens,  Picea  engelmannii,  Pi- 
nus sibirica,  and  others  (Schlatter  1935).  It  has  become 
evident,  however,  that  in  this  area  the  native  Pinus  cem- 
bra  and  Larix  decidua  are  the  most  successful  species. 

CONCLUDING  REMARKS 

The  Upper  Engadine  belongs  to  the  most  important 
natural  ranges  of  cembra  pine-larch  forests  in  the  Alps. 
However,  present  distribution,  composition,  and  struc- 
tures of  these  forests  can  only  be  explained  by  permanent 
utilization  that  began  in  prehistoric  time.  That  makes 
these  forests  much  different  from  almost  imtouched  high- 
altitude  forests  in  remote  areas  of  the  Rocky  Mountains, 
for  example. 

Since  tourism  has  become  the  main  economic  base,  while 
agriculture  and  pastoral  land  use  declined,  utilization  of 
the  forest  also  changed.  In  our  day  it  is  not  timber  produc- 
tion but  establishment  and  maintenance  of  a  vigorous 
shelter  forest  (protection  from  avalanches,  soil  erosion, 
torrential  washes,  etc.)  that  is  the  main  objective  of  forest 
management.  Locally,  grazing  of  the  high-altitude  forest 
by  cattle  and  game  may  run  counter  to  the  objective  and 
should  be  reduced  as  much  as  possible. 

However,  the  forest  also  is  of  great  importance  for  tour- 
ism. Thus,  larch,  which  has  a  high  esthetic  value  because 
of  its  bright  autvmin  colors,  is  favored  by  special  manage- 
ment practices  interrupting  natural  succession.  On  the 
other  hand,  lower  elevation  forest  is  still  used  for  produc- 
tion of  merchantable  timber.  Swiss  stone  pine  is  used  in 
particular  for  paneling,  carvings,  and  making  furniture  of 
typical  local  design. 


15 


PINUS  MUCO 

LARIX  DECIDUA 

9  <* 

PINUS  CEMBRA 


Figure  14 — Distribution  of  Pinus  mugo  and  AInus  viridis  on  the  northeast-exposed  slope  of  the  Bernina  vaiiey. 
Pinus  mugo  is  confined  to  relatively  dry  convex  topography,  while  AInus  viridis  grows  on  the  moist  sites  (from 
Holtmeier  1967). 


REFERENCES 

Auer,  Christian.  1947.  Untersuchungen  iiber  die  natiir- 
liche  Verjiingung  der  Larche  im  Arven-Larchenwald 
des  Oberengadins.  Mitteilungen  der  Schweizerischen 
Anstalt  fiir  das  forstliche  Versuchswesen.  25:  3-140. 

Campell,  Eduard.  1944.  Der  Wald  des  Oberengadins  im 
Wandel  der  Zeiten.  Festschrift  124.  Jahresversamm- 
lung  der  Schweizerischen  Naturforschenden 
Gesellschaft:  93-102. 

Conrad,  H.  1940.  Beitrag  zur  Frage  der  urgeschichtHchen 
Besiedlung  des  Engadins.  Jahresbericht  Historisch- 
Antiquarische  Gesellschaft  Graubixndens.  70. 
Jahresbericht:  1-40. 


Figure  15 — Birch  {Betula  pubescens)  growing 
in  an  avalanche  track. 


Ellenberg,  Heinz.  1978.  Vegetation  Mitteleuropas  mit  den 
Alpen  in  okologischer  Sicht.  2d  ed.  Stuttgart:  Verlag 
Eugen  Ulmer.  981  p. 

Heitz,  Christian.  1975.  Vegetationsentwicklung  und 
Waldgrenzschwankungen  des  Spatglazials  iind  Post- 
glazials  im  Oberhalbstein  (Graubiinden/Schweiz). 
Biindnerwald.  28(2):  45-52. 

Holtmeier,  Friedrich-Karl.  1965.  Die  Waldgrenze  im 
Oberengadin  in  ihrer  physiognomischen  und  okolo- 
gischen  DifFerenzierung.  University  of  Bonn,  FRG. 
163  p.  Dissertation. 

Holtmeier,  Friedrich-Karl.  1966a.  Die  Malojaschlange 
und  die  Verbreitung  der  Fichte.  Beobachtungen  zur 
Klimaokologie  des  Oberengadins.  Wetter  und  Leben. 
18:  105-108. 

Holtmeier,  Friedrich-Karl.  1966b.  Die  okologische  Funk- 
tion  des  Tannenhahers  im  Zirben-Larchenwald  und  an 
der  Waldgrenze  im  Oberengadin.  Journal  fur  Omi- 
thologie.  4:  253-260. 

Holtmeier,  Friedrich-Karl.  1967.  Die  Verbreitung  der 
Holzarten  im  Oberengadin  unter  dem  Einflui3  des 
Menschen  und  des  Lokalklimas.  Erdkunde.  21:  249-258. 

Holtmeier,  Friedrich-Karl.  1973.  Geoecological  aspects  of 
timberlines  in  Northern  and  Central  Europe.  Arctic  and 
Alpine  Research.  5(3):  45-54. 

Holtmeier,  Friedrich-Karl.  1974.  Geookologische  Beobach- 
tungen und  Studien  an  der  subarktischen  und  alpinen 
Waldgrenze  in  vergleichender  Sicht  (nordliches  Fenno- 
skandien/Zentralalpen).  Erdwissenschaftliche  For- 
schung.  8.  130  p. 

Holtmeier,  Friedrich-Karl.  1981.  What  does  the  term 
"Krummholz"  really  mean?  Observations  with  special 
reference  to  the  Alps  and  the  Colorado  Front  Range. 
Mountain  Research  and  Development.  1(3/4):  253-260. 

Holtmeier,  Friedrich-Karl.  1990.  Disturbance  and  man- 
agement in  larch-cembra  pine  forests  in  Europe.  In: 


16 


Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps.  Pro- 
ceedings— symposium  on  whitebark  pine  ecoystems: 
ecology  and  management  of  a  high-momitain  resoiirce. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agricultvire,  Forest  Service,  Intermoxantain  Research 
Station:  25-36. 

Holtmeier,  Friedrich-Karl.  1993.  European  larch  in 
Middle  Europe.  In:  Proceedings — symposium  on  ecology 
and  management  of  larix  forests:  a  look  ahead.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain  Research  Station.  [In  preparation]. 

Kleiber,  Helga.  1974.  Pollenanalytische  Untersuchvmgen 
zum  Eisriickgang  und  zur  Vegetationsgeschichte  im 
Oberengadin  I.  Botanisches  Jahrbuch  fiir  Systematik. 
94(1):  1-53. 


Mattes,  Hermann.  1978.  Der  Tannenhaher  im  Engadin — 

Studien  zu  seiner  okologischen  Funktion  im  Arvenwald. 

Miinstersche  Geographische  Arbeiten.  2.  87  p. 
Mattes,  Hermann.  1982.  Die  Lebensgemeinschaft  von 

Tannenhaher  und  Arve.  Eidgenossische  Anstalt  fur  das 

forstliche  Versuchswesen.  Berichte.  24.  74  p. 
Oberdorfer,  Erich.  1970.  Pflanzensoziologische  Exkur- 

sionsflora  fiir  Siiddeutschland.  3d  ed.  Stuttgart:  Verlag 

Eugen  Ulmer.  987  p. 
Schlatter,  A.  J.  1935.  Die  Aufforstungen  und  Verbau- 

xmgen  des  Oberengadins  in  den  Jahren  1875-1934. 

Schweizerische  Zeitschrift  fur  Forstwesen.  86:  309-328. 


17 


Evolution  and 
Taxonomy 


International  Workshop 
St.  Moritz  1 992 

S   


18 


GENETIC  DIFFERENTIATION  AND 
PHYLOGENY  OF  STONE  PINE 
SPECIES  BASED  ON  ISOZYME  LOCI 

Konstantin  V.  Krutovskii 
Dmitri  V.  Politov 
Yuri  P.  Altukhov 


Abstract— F-statistics  analysis  based  on  isozyme  loci  data  has 
revealed  comparatively  low  levels  of  intraspecific  genetic  differen- 
tiation among  natural  populations  of  Eurasian  stone  pine  species: 
Siberian  stone  pine  {Pinus  sibirica  Du  Tour),  Korean  stone  pine 
{Pinus  koraiensis  Siebold  et  Zucc),  and  Japanese  (dwarf  Siberian 
or  mountain)  stone  pine  {Pinus  pumila  [Pall.]  Kegel).  Only  about 
2  to  4  percent  of  the  total  isozyme  gene  diversity  was  due  to  inter- 
population  variation  (FgT  =  0.021-0.040),  and  the  overwhelming 
part  of  the  total  variation,  over  96  percent,  belonged  to  intrapopu- 
lation  variation.  These  data  are  discussed  in  relation  to  distin- 
guishing features  of  conifer  biology,  in  general,  and  to  a  specific 
mode  of  stone  pine  seed  distribution  by  nutcrackers,  in  particular. 
Cluster  analysis  of  more-studied  Siberian  stone  pine  populations 
has  shown  that  genetic  distances  based  on  isozyme  allele  frequen- 
cies reflect  geographical  distribution  of  these  populations  in  spite 
of  the  generally  low  values  of  genetic  distances  (D  =  0.001-0.031, 
on  the  average  of  0.006). 

The  authors  have  studied  phylogenetic  relationships  between  all 
stone  pine  species  (subsection  Cembrae),  including  one  European 
stone  pine  population— Swiss  stone  pine  {Pinus  cembra  L.)  and  one 
population  of  the  only  representative  of  stone  pine  species  in  North 
America— whitebark  pine  {Pinus  albicaulis  Engelm.),  using  genetic 
distances  based  on  16  isozjmie  loci,  whose  genetic  control  has  been 
resolved  for  all  stone  pine  species.  The  results  have  confirmed 
stone  pine  {Cembrae)  species  to  be  a  compact  group  of  very  closely 
related  and  genetically  similar  species,  having  monophyletic  ori- 
gin, supposedly  from  ancient  Siberian  stone  pine.  The  smallest 
distance  (D  =  0.065)  has  been  found  between  Pinus  cembra  and 
Pinus  sibirica  that  corresponded,  mostly,  to  their  subspecies  taxo- 
nomic  status.  Other  interspecific  distances  were  in  good  agree- 
ment with  the  species  status  (D  =  0.121-0.268). 

As  to  the  controversy  on  the  taxonomy  of  whitebark  pine,  this 
species  undoubtedly  belongs  to  subsection  Cembrae  (D  =  0.121- 
0.256).  The  results  of  phylogenetic  study  have  been  confirmed  by 
the  recently  obtained  data  of  cpDNA  restriction  fragment  analyses. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  ICnowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Konstantin  V.  Krutovskii  is  Senior  Research  Scientist,  Laboratory  of 
Population  Genetics,  N.I.  Vavilov  Institute  of  General  Genetics,  Russian 
Academy  of  Sciences,  GSP-1  Moscow  117809  B-333,  Russia  (current  ad- 
dress until  1995:  Department  of  Forest  Genetics  and  Tree  Improvement, 
Georg-August-University,  Biisgenweg  2,  Gottingen,  37077,  Germany); 
Dmitri  V.  Politov  is  Research  Scientist,  same  laboratory;  Yuri  P.  Altukhov 
is  Professor,  Member  of  Russian  Academy  of  Sciences,  Head  of  the  Labora- 
tory of  Population  Genetics,  Director  of  the  Institute  of  General  Genetics. 


Five  stone  pine  species  are  traditionally  united  in  sub- 
section Cembrae  of  section  Strobus,  subgenus  Strobus, 
genus  Pinus  (Critchfield  and  Little  1966;  Shaw  1914). 
European  or  Swiss  stone  pine  {Pinus  cembra)  is  distrib- 
uted in  the  Alps  and  Carpathians  and  has  the  smallest 
area  among  stone  pine  species.  An  east  Asian  stone  pine 
species,  Korean  pine  (P.  koraiensis),  is  widely  distributed 
in  eastern  Russia  (Russian  "Far  East"),  Manchuria,  and 
Japan,  but  mostly  in  Korea.  Two  other  stone  pine  species 
are:  Japanese  stone  pine  {P.  pumila),  widely  distributed 
in  northeastern  Siberia  extending  to  eastern  Asia,  as  far 
as  central  Japan;  and  Siberian  stone  pine  (P.  sibirica), 
widely  distributed,  mostly  in  Siberia  extending  to  northern 
Mongolia  and  to  the  eastern  border  of  Europe.  Whitebark 
pine  (P.  albicaulis)  is  widely  distributed  in  western  North 
America,  extending  from  British  Coliunbia  to  the  southern 
Sierra  Nevada,  and  represents  the  only  stone  pine  species 
growing  outside  Eurasia. 

All  these  species  are  of  great  interest  for  different  kinds 
of  investigations,  but  some  main  aspects  make  these  species 
very  attractive  from  evolutionary  and  population  genetics 
points  of  view  and  explain  the  necessity  of  their  genetic 
study. 

First,  they  occupy  a  vast  territory  in  Eurasia  and  North 
America  and  most  of  them  belong  to  the  main  forest-forming 
species.  Thus,  they  have  great  ecological  and  economic  sig- 
nificance, especially  in  Eurasia.  They  are  subjected  to  inten- 
sive harvesting  and,  consequently,  need  special  breeding 
and  gene  conservation  programs  based  on  genetic  data. 

Second,  stone  pine  species  consist  of  different  kinds  of  pop- 
ulations. For  example,  Swiss  stone  pine  includes  mostly 
small,  isolated  populations  and  occupies  a  relatively  small 
territory.  On  the  other  hand,  Siberian  stone  pine  includes 
huge,  continuous  populations  occupying  a  very  leirge  terri- 
tory in  Siberia.  Some  of  the  stone  pine  species  are  com- 
pletely isolated  from  others:  Swiss  stone  pine,  growing  in 
Europe,  and  whitebark  pine,  growing  in  North  America, 
stand  apart  from  Asian  stone  pine  species.  At  the  same 
time,  Siberian,  Korean,  and  Japanese  stone  pines  have 
sympatric  zones.  Thus,  from  evolutionary  and  population 
genetics  points  of  view  and  for  gene  conservation  programs 
it  would  be  interesting  and  important  to  study  genetic  dif- 
ferentiation, gene  dispersal  patterns,  and  mating  system 
and  population  genetic  structure  of  these  species  and  com- 
pare them  with  each  other. 

Third,  in  spite  of  the  fact  that  stone  pines  have  a  wide 
geographical  distribution  and  some  morphological  and 
ecological  differences,  they  are  grouped  into  a  separate 
subsection  Cembrae,  because  they  have  some  diagnostic 


19 


morphological  traits  that  distinguish  them  from  other 
Strobus  species.  Most  significant  of  these  are  wingless  large 
seeds  and  indehiscent  cones  that  remain  closed  in  matu- 
rity. These  traits  are  supposed  to  be  the  result  of  long  co- 
evolution  of  stone  pines  and  nutcrackers  (Nucifraga  ssp.), 
who  play  a  crucial  role  in  natural  reproduction  and  dis- 
tribution of  stone  pine  species  (Lanner  1980,  1982,  1990; 
Tomback  1983;  Tomback  and  Linhart  1990;  Tomback  and 
others  1990).  Nevertheless,  taxonomy  of  these  species  is 
still  being  discussed.  Some  specialists  consider  Swiss  stone 
pine  a  subspecies  of  Siberian  stone  pine  (or  vice  versa)  and 
some  think  that  whitebark  pine  should  be  included  into 
subsection  Strobi  because  of  its  morphological  and  ecologi- 
cal resemblance  to  some  Strobi  species  (Critchfield  1986; 
Millar  and  Kinloch  1991).  Taking  all  this  into  account,  the 
use  of  genetic  markers  for  study  of  phylogenetic  relation- 
ships between  stone  pine  species  would  be  very  appropriate 
and  fruitful. 

Thus,  the  main  tasks  of  our  research  were  estimation  of 
genetic  differentiation  and  subdivision  of  stone  pine  popula- 
tions and  study  of  phylogenetic  relationships  between  stone 
pine  species  using  isozyme  loci  as  genetic  markers. 

The  advantages  and  disadvantages  of  using  isozymes  in 
plant  systematics  and  phylogeny  have  been  fairly  thoroughly 
discussed  in  scientific  literature  (see  for  example,  Crawford 
1983;  Gottheb  1977;  Strauss  and  others  1992).  We  would 
only  like  to  note  here  such  taxonomically  most  important 
advantages  as,  first,  the  possibility  of  using  the  same  trait 
(isozyme  electrophoretic  pattern  encoded  by  homologous 
loci)  to  classify  taxons  of  almost  any  phylogenetic  level,  both 
related  and  unrelated,  and  second,  the  possibility  of  quan- 
titative estimation  of  divergence  degree.  One  of  the  prob- 
lems of  traditional  systematics  is  associated  with  the  objec- 
tive choice  of  diagnostic  traits,  but  initial  selection  of  such 
traits  usually  leads  to  their  preliminary  "weighting"  (evalu- 
ation) and  inevitably  produces  subjectivity.  It  is,  therefore, 
extremely  important  to  use,  above  all,  such  traits  in  taxo- 
nomic  studies  as  can  be  taken  for  analysis  without  "weigh- 
ing," more  or  less  arbitrarily,  and  about  whose  variability 
nothing  has  been  known  a  priori.  Molecular  genetic  mark- 
ers (isozyme  loci  and  DNA  markers)  are  one  of  the  best, 
from  these  points  of  view. 

The  study  presented  here  is  the  first  one  devoted  to  study 
of  phylogeny  of  the  whole  subsection  Cembrae  using  isozyme 
markers  and  is  only  a  part  of  long-term  studies  on  popula- 
tion genetics  of  stone  pine  species  carried  out  in  the  Labora- 
tory of  Population  Genetics  of  N.  I.  Vavilov  Institute  of  Gen- 
eral Genetics  (Russian  Academy  of  Sciences,  Moscow,  Russia) 
on  the  basis  of  isozyme  analysis.  More  details  on  inheritance 
and  linkage  of  isozjnne  loci,  mating  system,  population  ge- 
netic structure,  and  other  aspects  of  stone  pine  population 
genetics  can  be  found  in  our  previously  published  papers 
(Krutovskii  and  Politov  1992;  Krutovskii  and  others  1987, 
1988,  1989,  1990;  PoHtov  1989;  Politov  and  Krutovskii  1990; 
Politov  and  others  1989,  1992)  and  in  the  accompanying  pa- 
per included  in  these  proceedings  by  Politov  and  Krutovskii. 

MATERIAL  AND  SAMPLING  SITES 

Seventeen  single-tree  and  two  bulk  seed  samples  from 
natural  populations  have  been  used  for  isozyme  analysis  in 
our  study.  Eleven  localities  of  Siberian  stone  pine,  three 


localities  each  of  Korean  and  Japanese  stone  pines,  and 
one  locality  each  of  Swiss  stone  and  whitebark  pines  have 
been  studied  on  16-20  isoz3niie  loci  (table  1).  More  details 
about  seed  samples  and  characteristics  of  the  sampled 
natural  populations  can  be  found  in  our  other  publications 
(Krutovskii  and  Politov  1992;  Krutovskii  and  others  1988, 
1989,  1990;  Pohtov  1989;  Pohtov  and  Krutovskii  1993; 
Pohtov  and  others  1992). 

ELECTROPHORETIC  ANALYSIS 

Isozymes  of  seeds  have  been  studied  by  electrophoresis 
in  starch  gel.  Method  of  electrophoretic  analysis,  specimen 
preparation,  the  buffer  systems  used,  genetic  interpretation 
of  zymograms,  designations  of  allozymes,  alleles,  and  loci, 
and  inheritance  of  enzyme  systems  have  been  described 
elsewhere  (Krutovskii  and  others  1987;  Politov  1989).  Geno- 
types of  trees  have  been  determined  by  analyzing  segrega- 
tion of  allozyme  alleles  among  their  endosperms,  which  are 
haploid  in  conifers  and  genetically  identical  to  maternal 
tree  gametes.  Together  with  adult  tree  allele  frequencies, 
pollen  allele  frequencies  inferred  from  effective  pollen  gene 
pool  (paternal  alleles  of  the  analyzed  seed  embryos)  have 
been  used  for  estimation  of  genetic  distances.  These  fre- 
quencies of  successful  pollen  give  probably  more  represen- 
tative estimation  of  the  studied  populations'  gene  pools  due 
to  predominantly  outcrossing  pollination  among  stone  pine 
species  (Pohtov  and  Krutovskii  1990,  1993). 

In  total,  536  trees  fi-om  17  localities  (11-75  trees  each)  and 
two  bulk  seed  samples  of  five  stone  pine  species  have  been 
analyzed  on  16-20  isozyme  loci.  More  than  3,500  (7-20  per 
tree)  embryos  have  been  analyzed  on  a  smaller  amount  of 
isoz5Tne  loci  because  of  poor  resolution  of  some  enzyme  sys- 
tems in  embryo  tissue.  A  relatively  small  number  of  trees 
have  been  analyzed  in  some  populations  and  species,  but 
according  to  some  theoretical  (Nei  1987a,b)  and  experimen- 
tal investigations  (Shurkhal  and  others  1992),  the  number 
of  loci  used  in  interspecific  phylogenetic  analysis  is  more 
important  than  the  sample  size.  We  have  studied  phyloge- 
netic relationships  between  all  stone  pine  species  using  ge- 
netic distances  based  on  16  isozyme  loci,  Adh-1,  Dia-2,  Fe-2, 
Gdh,  (iot-1,  Got-2,  Got-3,  Idh,  Mdh-1,  Mdh-2,  Mdh-3,  Mdh-4, 
Pgi-2,  Pgm-1,  Pgm-2,  and  Skdh-1,  whose  allelic  variation  has 
been  unequivocally  determined  for  all  stone  pine  species. 
Such  a  number  of  loci  is  used  commonly  and  gives  quite 
reasonable  phylogenetic  trees  (Nei  1987a,b). 

DATA  ANALYSIS 

Determination  of  parameters  of  intra-  and  interpopula- 
tional  genetic  diversity,  estimation  of  genetic  differentia- 
tion and  genetic  distances,  and  clustering  and  construction 
of  dendrograms  for  stone  pine  populations  have  been  car- 
ried out  by  using  IBM  PC  version  1.7  of  the  computer  pro- 
gram BIOSYS-1  (Swofford  and  Selander  1981)  and  IBM 
PC-program  RESTSITE,  version  1.1  (Nei  and  Miller  1990). 

Using  our  data,  we  have  estimated  different  kinds  of  ge- 
netic distances  and  tried  different  methods  of  clustering. 
Methods  for  calculating  genetic  distances  and  methods  of 
clustering  differ  in  their  assumptions  concerning  homoge- 
neity of  evolutionary  rates  or  proportionality  to  the  time  of 
divergence.  Consequently,  topology  of  derived  phylogenetic 


20 


Table  1 — Name,  abbreviated  designation,  sample  size,  and  geographic  location  of  sampling  sites 


Locality 


Designation 


Number 
of  trees 


Region 


Siberian  Stone  Pine 


Filin  Klyuch 


Mutnaya  Rechka  M 

Maiyi  Kebezh  MK 

Sobach'ya  Rechka  SR 

Listvyanka  L 

Yailyu  YA 

Smokotino  S 

Zorkartsevo-1  Z-1 

Zorkartsevo-3  Z-3 

Turukhansk  TU 


Nyalino  NYA 


Ust'-Chornaya  UCH 


Grustnyi  GR 
Uttyveem  UT 
Kamenskoye  K 


Malokhekhtsirskoye  MKH 
Khor  KHO 
Sikhote-Alin'  SA 


Livingstone  Falls  LF 


53 


75 
15 
12 
41 
43 

43 
34 
34 

MOO 


Western  Sayan  Mountains  (Ermakovskii 
District,  Krasnoyarsk  Territory, 
eastern  Siberia) 


Altai  Mountains  (Gorno-Altaiskaya 

Autonomous  Region,  eastern  Siberia) 
Tomskaya  Region  (western  Siberia) 


'88 


North  of  Krasnoyarsk  Territory, 
(Turukhanskii  District,  North  of 
eastern  Siberia) 
Khanty-Mansi  Autonomous  District 
(Tyumenskaya  Region,  North  of 
western  Siberia) 

Swiss  Stone  Pine 

15  Eastern  Carpathians  (western  Ukraine) 

Japanese  Stone  Pine 

54  North  of  Kamchatka  Peninsula 

55 

55 


Korean  Stone  Pine 


30 
16 
11 

Whitebark  Pine 

52 


Khabarovsk  Territory 

(Russian  "Far  East") 

Primorskii  Territory  (Russian  "Far  East") 


Rocky  Mountains  Forest  Reserve, 
Bow-Crow  Forest  (southwestern 
Alberta,  Canada) 


'Number  of  analyzed  seeds  from  mixed  sample  obtained  from  large  number  (>100)  of  trees. 


trees  based  on  different  distances  and  methods  of  cluster- 
ing can  vary  significantly  for  the  same  data. 

Among  methods  of  clustering  that  assimie  homogeneity 
of  evolutionary  rates,  an  unweighted  pair  group  method 
using  arithmetic  averages,  UPGMA,  is  the  most  widely 
used  one  (Sneath  and  Sokal  1973).  Among  methods  that 
do  not  assume  homogeneity  of  evolutionary  rates,  the  dis- 
tance Wagner  procedure  (Farris  1972)  and  neighbor-joining 
method,  NJM  (Saitou  and  Nei  1987)  are  the  most  widely 
used  ones. 

The  most  common  genetic  distances  used  in  isozjnne 
studies  of  trees  are  Nei's  (Nei  1972,  1978),  Rogers'  (1972), 
and  CavalU-Sforza  and  Edwards'  (1967)  distances.  How- 
ever, Nei's  distance,  D,  along  with  advantages  has  some 
disadvantages,  the  main  one  being  that  it  is  not  metric 
because  it  does  not  obey  the  triangle  inequality  and,  conse- 
quently, cannot  be  used  in  the  distance  Wagner  procedure 
or  NJM.  Thus,  we  have  used  D  (Nei  1972)  only  in  UPGMA 
dendrograms. 


Taking  into  accoimt:  (1)  that  homogeneity  of  evolution- 
ary (or  allozyme  allele  substitution)  rates  can  usually  be 
safely  assumed  in  comparing  conspecific  populations  of 
widespread,  continuous  tree  species  with  high  gene  flow, 
like  stone  pine  populations;  (2)  shortage  of  space  for  pre- 
sentation of  all  obtained  dendrograms;  and  (3)  possibility 
for  wider  comparison  vnth  published  data,  we  will  present 
only  UPGMA  dendrograms  based  on  Nei's  distance,  D, 
for  analysis  of  intraspecies  stone  pine  population  relation- 
ships. However,  to  avoid  problems  connected  with  possible 
nonhomogeneity  of  evolutionary  rates  of  different  species, 
we  will  present  NJM  dendrograms  of  stone  pine  species 
based  on  Cavalli-Sforza  and  Edwards'  (1967)  chord  dis- 
tances, Dch,  in  addition  to  UPGMA  dendrograms  based  on 
D  and  D.^. 

Nei's  genetic  distances  matrix  has  also  been  used  for 
principal  coordinate  analysis  of  Siberian  stone  pine  popula- 
tions using  the  computer  program  NTSYS-pc  (Rohlf  1988), 


21 


INTRASPECIFIC  DIFFERENTIATION 

Levels  of  Genetic  Differentiation  Between 
Populations — Distribution  of  intraspecific  genetic  variation 
between  populations  of  stone  pines  has  been  analyzed  for 
three  species  studied  in  several  localities  using  F-statistics 
(Nei  1977).  Levels  of  genetic  differentiation  among  popula- 
tions were  comparatively  low — only  about  2-4  percent  of 
the  total  intraspecific  isozyme  gene  variation  was  due  to 
interpopulation  variation  (FgT  =  0.021-0.040),  and  the  over- 
whelming part  of  the  total  variation,  over  96  percent,  be- 
longed to  intrapopulation  variation  (table  2).  Loci  Fe-2, 
Mdh-3,  and  Skdh-1  of  Siberian  stone  pine,  Adh-1,  Adh-2, 
Gdh,  Got-2,  Mdh-2,  Mdh-3,  Mdh-4,  and  Skdh-2  of  Japanese 
stone  pine,  and  Mdh-4  of  Korean  stone  pine  make  a  most 
significant  contribution  in  intraspecific  differentiation  of 
these  species  (table  2).  The  geographic  variation  of  these 
loci  may  be  adaptive  and  should  be  studied  in  detail  in  fu- 
ture investigations. 

Nei's  genetic  distances  between  populations  were  also 
small,  D  =  0.001-0.031  and  0.012,  on  the  average  (table  3). 
These  values  correspond  perfectly  to  the  calculations  car- 
ried out  using  references  on  22  pine  species  (Altukhov  and 
others  1989).  The  genetic  distances  between  closely  located 
(<20  km)  populations  of  seven  pine  species  equaled  0.009, 
on  the  average,  and  between  widely  dispersed  populations 
of  15  pine  species  reviewed — D  =  0.036,  on  average. 

Low  levels  of  interpopulation  differentiation  of  allozyme 
loci  are  usual  for  conifers.  Such  typical  factors  for  most  coni- 
fer populations  as  outcrossing,  wind  pollination,  seed  dis- 
persal by  wind  (most  conifers)  or  by  birds  (stone  pines,  some 
white  pines,  and  pinyons),  wide  continuous  ranges,  high 
population  density,  and  effective  size  are  considered  to  re- 
duce the  influence  of  genetic  drift  and,  therefore,  decrease 


heterogeneity  of  allele  frequencies  and  interpopulation 
genetic  differentiation  (Hamrick  1983;  Hamrick  and  Godt 
1989;  Hamrick  and  Loveless  1986;  Hamrick  and  others 
1981;  Loveless  and  Hamrick  1984). 

We  have  also  simmiarized  data  on  interpopulation  genetic 
differentiation  of  28  conifer  species  of  eight  genera  estimated 
using  FgT  or  analogous  GgT  parameters  of  differentiation 
(Politov  and  others  1992).  In  spite  of  a,  generally,  compar- 
atively low  differentiation  among  populations,  we  have 
found  a  quite  pronounced  tendency  to  a  significantly  smaller 
differentiation  (Fst  =  0.02-0.08)  among  widespread  species 
with  continuous  population  ranges  than  among  species  with 
interrupted  ranges  and  small,  and  often  isolated,  popula- 
tions (Fg-r  =  0.11-0.13).  Moreover,  according  to  our  review, 
samples  from  the  same  population  (samples  fi'om  localities 
within  several  kilometers  radius)  revealed,  commonly,  an 
even  much  lower  level  of  differentiation  (Fg^  =  0.006-0.02). 

The  three  stone  pine  species  studied  belong  to  the  first 
group  of  species  with  mainly  continuous  population  range. 
Besides,  four  of  the  11  Siberian  stone  pine  samples,  F,  M, 
MK,  and  SR,  have  been  collected  within  20  km  radius  from 
obviously  the  same  population  of  Malyi  Kebezh  River  Basin 
(eastern  Siberia,  table  1);  and  the  other  two  samples  have 
been  taken  fi^om  two  stands  near  Zorkal'tsevo  village  (west- 
em  Siberia),  Z-1  and  Z-3,  which  are  separated  from  each 
other  by  less  than  3  km.  The  other  localities  are  rather  re- 
mote from  one  another,  but  all  of  them  represent  only  the 
central,  continuous  part  of  the  species  area.  There  are  no 
significant  barriers  to  gene  flow  in  this  area. 

The  low  differentiation  observed  in  our  study  may  also 
be  due  to  the  limited  number  of  samples  taken,  mostly, 
from  more  or  less  optimal  zones  of  the  stone  pines  area, 
impeding  the  diversifying  selection  action  and  promoting 
the  balancing  one  (Altukhov  1990). 


Table  2 — The  estimates  of  F-statistics  for  populations  of  three  stone  pine  species  and  for  all  five  species 


Locus 

Siberian  stone  pine 

Korean  stone  pine 

Japanese  stone  pine 

All 

Ft 

Ft 

Fst 

F,s 

Ft 

Fst 

Fst 

Adh-1 

-0.030 

-0.006 

0.023 

0.078 

0.120 

0.045 

0.306 

0.321  ' 

0.022 

'0.563 

Adh-2 

{') 

{') 

{') 

(') 

{') 

-.068 

-.046 

\020 

Dia-1 

-.032 

-.007 

.024 

(') 

(') 

{') 

n 

{') 

Dia-2 

-.060 

-.053 

.006 

{') 

{') 

(') 

-.038 

-.013 

\025 

'.028 

Fe-2 

-.061 

-.012 

\046 

-.067 

-.021 

.043 

-.105 

-.097 

.007 

'.190 

Gdh 

-.012 

-.001 

.010 

-.098 

-.098 

.000 

.084 

.140 

^061 

'.637 

Got-1 

-.013 

-.001 

.012 

{') 

{') 

{') 

{') 

{') 

{') 

.002 

Got-2 

-.012 

-.001 

.010 

{') 

e) 

-.050 

-.022 

\027 

'.016 

Got-3 

(') 

{') 

{') 

.238 

.241 

.004 

.014 

.015 

.001 

'.596 

Lap-2 

.015 

.025 

.009 

-.071 

-.023 

.045 

-.048 

-.047 

.001 

Lap-3 

.066 

.079 

.013 

(') 

{') 

{') 

-.064 

-.050 

.013 

Mdh-2 

.214 

.231 

.023 

-.251 

-.225 

.020 

-.004 

.015 

^019 

'.240 

Mdh-3 

-.049 

.000 

\047 

-.067 

-.021 

.043 

-.051 

.084 

M28 

'.118 

Mdh-4 

-.068 

-.055 

.012 

.017 

.104 

'.089 

-.021 

-.013 

\ooa 

'.362 

Pgi-2 

.114 

.123 

.010 

e) 

{') 

{') 

-.111 

-.107 

.003 

'.407 

Pgm-1 

-.074 

.051 

.022 

-.193 

-.160 

.028 

.177 

.189 

.016 

'.187 

Pgm-2 

{') 

(') 

{') 

{') 

?) 

-.014 

-.013 

.002 

'.046 

Skdh-1 

-.146 

-.118 

\025 

-.316 

-.194 

.092 

-.053 

-.051 

.002 

'.187 

Skdh-2 

{') 

{') 

{') 

{') 

{') 

.099 

.163 

'.071 

Mean 

-.043 

-.018 

.025 

-.077 

-.033 

.040 

-.002 

.020 

.021 

.396 

'This  locus  has  statistically  significant  heterogeneity  of  allele  frequencies  among  populations. 

n"he  genetic  base  of  isozyme  variation  of  this  locus  is  not  determined  or  allelic  variation  cannot  be  compared  with  other  species, 
^his  locus  is  monomorphic  in  this  species. 


22 


Table  3 — Genetic  distances,  D,  (Nei  1972)  for  five  stone  pine  species 


Sibarian 

Ivor  oan 

Species 

stone  pine 

stone  pine 

stone  pine 

Stone  pine 

Siberian  stone  pine 

11 

0.006 
(0.002-0.016) 

Swii^^  stonp  oinp 

"1 

0  065 
(0.055-0.077) 

Jaoanese  stone  oine 

3 

0.202 
(0.176-0.253) 

0.233 
(0.215-0.265) 

0  010 
(0.003-0.014) 

Korean  stone  pine 

3 

0.235 

0.268 

0.105 

0.009 

(0.207-0.261) 

(0.264-0.275) 

(0.096-0.119) 

(0.004-0.014) 

Whitebark  pine 

1 

0.121 

0.130 

0.185 

0.256 

(0.111-0.129) 

(0.166-0.220) 

(0.253-0.261) 

'N  =  number  of  localities  studied. 


And  finally,  the  low  differentiation  between  stone  pine 
populations  can  be  considered  the  result  of  a  specific  mode 
of  seed  dispersal  by  nutcrackers  able  to  carry  their  seeds 
over  very  long  distances  (up  to  20  km,  according  to  some 
observations).  The  efficiency  of  this  mode  and  its  influence 
on  the  population  structure  were  repeatedly  emphasized 
by  researchers  studying  relationships  between  nutcrackers 
and  whitebark  pine,  Swiss  stone  pine,  limber  pine  {Pinus 
flexilis  James),  and  pinyons  (subsection  Cembroides) 
(Carsey  and  Tomback  1992;  Furnier  and  others  1987; 
Lanner  1980,  1990;  Schuster  and  Mitton  1991;  Tomback 
and  others  1990,  1992). 

Cluster  and  Principal  Coordinates  Analysis — 

Despite  the  low  values  of  genetic  distances,  UPGMA  cluster- 
ing of  stone  pine  samples  based  on  these  distances  reflects, 
in  general,  their  geographical  origin.  More  detailed  analy- 
ses have  been  performed  on  Siberian  stone  pine,  which  has 
been  studied  more  than  other  stone  pines.  We  can  tenta- 
tively divide  all  11  Siberian  stone  pine  localities  into  three 
regions,  according  to  their  ecology  and  geographic  location. 
The  first  of  them  can  be  called  "Southern  Siberia  Moimtains" 
and  includes  one  population  from  the  Altai  Mountains  (Ya) 
and  five  localities  in  the  western  Sayan  (FK,  M,  MK,  SR, 
and  L);  the  second  can  be  called  "Western  Siberia  Plain" 
and  includes  populations  from  Tomsk  region  (S,  Z-1,  and 
Z-3);  the  third,  "Northern  Siberia,"  includes  populations 
from  Khanty-Mansi  Autonomous  District  (NYa)  and  from 
the  north  of  Krasnoyarsk  Territory  (TU).  The  dendrogram 
obtained  corresponds  exactly  to  this  classification  (fig.  1). 
These  three  regions  form  three  different  genetic  groups 
(clusters)  of  populations. 

Korean  pine  samples  also  form  clusters  in  accordance 
with  their  population  localization  (two  geographically  more 
closely  spaced  stands,  MKh  and  Kho,  have  turned  out  to 
be  genetically  more  similar).  The  same  is  true  of  Japanese 
stone  pine  populations  (fig.  1).  Principal  coordinates  analy- 
ses based  on  the  matrix  of  Nei's  genetic  distances  between 
Siberian  stone  pine  populations  also  have  been  performed. 
This  multidimensional  analysis  has  produced  almost  the 
same  result  as  the  cluster  analysis.  Genetic  relationships 
between  Siberian  stone  pine  populations,  graphically  pre- 
sented in  two-dimensional  space  of  the  first  two  principal 
coordinates,  have  perfectly  reflected  the  real  geographical 


distribution  of  these  populations  (fig.  2).  Principal  coordi- 
nates and  cluster  analyses  complement  each  other  and  have 
produced  almost  the  same  result  in  our  case,  but  the  first, 
probably,  gives  more  information  about  population  genetic 
relationships  due  to  better  data  presentation.  Multivariate 
analysis  techniques  are  supposed  to  provide,  basically,  bet- 
ter possibility  for  revealing  genetic  differentiation  in  coni- 
fers, as  compared  to  univariate  methods  (Krutovskii  and 
Politov  1992;  Yeh  and  others  1985). 

Statistically  significant  positive  correlation  between  geo- 
graphical and  genetic  distances  has  been  found  for  Siberian 
stone  pine  populations  (r  =  0.844,  P  <  0.001).  Such  correla- 
tion was,  however,  absent  for  samples  from  closely  spaced 
localities  of  the  Malyi  Kebezh  River  Basin.  The  limited 
number  of  samples  studied  has  not  allowed  such  analysis 
for  other  stone  pine  species,  but  analogous  results  have 


0.010 


0.005 
I 


D 

F 

L 
M 
MK 

Southern 

Siberia 

Mountains 

YA 

SR  _ 

Z-1 

Z-3 

Western 

Siberia 

Plain 

NYA  _ 

Northern 
Siberia 

MKH~ 
KHO_ 

1  Khabarovsk" 
Territory 

SA 

Primorski!  - 
Territory 

GR 
UT 

K  - 

Kamchatka 
Peninsula 

Siberian 

stone  pine 


Korean 
stone  pine 

Japanese 
stone  pine 


Figure  1 — UPGMA  phenograms  of  stone  pine 
populations  based  on  Nei's  genetic  distances, 
D,  (Nei  1972).  (See  full  names  and  origins  of 
populations  in  table  1 .) 


23 


NYA 


Southern 
Siberia 
Mountains 


Western 
Siberia  Plain 


-0.15 


Figure  2 — Projections  of  Siberian  stone  pine 
populations  in  two-dimensional  space  of  the  first 
two  principal  coordinates,  X  and  Y,  according  to 
Nei's  genetic  distances,  D,  (Nei  1972).  (See  full 
names  and  origins  of  populations  in  table  1 .) 


also  been  obtained  for  some  other  pine  species  (Furnier 
and  Adams  1986;  Steinhoff  and  others  1983;  Wheeler  and 
Guries  1982),  including,  probably,  that  isolation  by  distance 
(Nei  1975),  introgressive  hybridization  (Millar  1983),  or 
gradiently  changing  selection  (Bergmann  1978)  could  play 
a  significant  role  in  genetic  differentiation  of  some  pine 
species. 

We  have  also  analyzed  correlation  between  geographical 
and  Nei's  distances  calculated  for  Siberian  stone  pine  popu- 
lations using  data  on  individual  loci  (allele  frequencies  of 
eight  loci  in  the  effective  pollen  pool  have  been  used).  For 
loci  Adh-1,  Dia-2,  Fe-2,  Pgm-1,  and  Skdh-1  correlation  was 
significant.  The  other  loci  (Lap-3,  Mdh-2,  and  Pgi-2)  dis- 
played lack  of  correlation.  This  may  suggest  a  selective  role 
of  certain  loci.  Patterns  of  spatial  differentiations  seem  to 
be  determined  by  complex  interaction  of  gene  flow  and  selec- 
tion. Distribution  of  genetic  variability  may  be  also  affected 
by  "evolutionary  footprints,"  such  as,  for  example,  Siberian 
population  descent  from  different  refugiums. 

INTERSPECIFIC  DIFFERENTIATION 

Genetic  Distances  Between  Stone  Pine  Species — 

We  have  estimated  both  genetic  distances,  D^j,  and  D,  be- 
tween all  stone  pine  species  based  on  16  isozyme  loci  data, 
but  only  D  values  are  presented  in  table  3,  because  of  space 
limits.  The  smallest  genetic  distances  have  been  found 
between  Siberian  and  Swiss  stone  pines  (D  =  0.052-0.077). 
They  are  just  a  little  larger  than  those  between  subspecies 
of  lodgepole  pine  (Pinus  contorta  Dougl.  ex  Loud.),  D  =  0.008- 
0.019,  approximately  0.012  on  the  average  (Wheeler  and 
others  1983),  similar  to  the  distances  between  subspecies 
of  brutian  pine  {Pinus  brutia  Ten.),  D  =  0.03-0.09  (Conkle 
and  others  1988),  and  a  little  smaller  than  distances  be- 
tween two  very  closely  related  species,  jack  pine  (Pinus 
banksiana  Lamb.)  and  lodgepole  pine,  which  have  a  wide 
zone  of  introgressive  hybridization,  D  =  0.097,  according  to 


the  study  by  Dancik  and  Yeh  (1983),  or  D  =  0.108,  according 
to  Wheeler  and  Guries  (1987).  Nei's  distances  estimated 
between  other  pairs  of  stone  pine  species  (D  =  0.121-0.268, 
table  3)  are  close  to  the  values  reported  for  pine  species 
of  subsections  Contortae:  0.078-0.158  (Wheeler  and  others 
1983)  and  Oocarpae:  0.11-0.27  (Millar  and  others  1988). 
Approximately  the  same  value,  D  =  0.27-0.43,  has  been 
reported  for  Pinus  brutia-Pinus  halepensis  comparison 
(Conkle  and  others  1988).  We  have  estimated  the  average 
genetic  distance  between  species  fi-om  the  same  subsections 
using  a  few  references  (Conkle  and  others  1988;  Millar  and 
others  1988;  and  some  others).  It  equaled  0.185,  on  the  av- 
erage, for  14  pine  species  and  surprisingly  corresponded  to 
the  average  distance  estimated  for  all  stone  pine  species  in 
our  study — D  =  0.180,  and  for  four  Eurasian  stone  pine  spe- 
cies studied  by  Shurkhal  and  others  (1991a,b)— D  =  0.138 
(for  comparison,  D  =  0.185,  on  the  average,  for  four  Eurasian 
stone  pines  used  in  our  study).  The  small  discrepancy  be- 
tween the  latter  two  estimates  is  not  surprising,  taking  into 
account  the  sampling  strategy  used  by  Shurkhal  and  others 
(1991a,b;  1992).  They  have  analyzed  only  a  few  trees  of 
each  species  (mostly,  from  Botanical  Garden  collections), 
which  should  inevitably  bias  genetic  distance  estimates  to 
smaller  values,  due  to  a  higher  probability  of  random  fixa- 
tion of  most  common,  frequent  alleles  in  small  samples  of 
trees  from  different  species. 

Levels  of  interspecific  genetic  differentiation  obtained 
between  stone  pines  strongly  confirm  their  taxonomical 
status  as  closely  related  species  of  the  same  subsection. 
Nei's  genetic  distances  obtained  for  species  belonging  to 
different  subsections  or  sections  are  usually  higher  than 
0.3  and  vary  between  0.3  and  0.8  (Shurkhal  and  others 
1991a,b;  1992).  According  to  phylogenetic  analysis  carried 
out  by  Shiu-khal  and  others  (1991a,b;  1992)  using  isoz)Tne 
loci,  fotir  stone  pine  species  of  subsection  Cembrae  (white- 
bark  pine  has  not  been  analyzed)  and  three  white  pine  spe- 
cies of  subsection  Strobi,  Balkan  or  Macedonian  pine  (Pinus 
peuce  Griseb.),  blue  or  Himalayan  pine  (Pinus  griffithii 
McCleland),  and  eastern  white  or  We3maouth  pine  (Pinus 
strobus  L.),  form  separate  clusters.  The  Nei's  genetic  dis- 
tance between  these  two  very  closely  related  subsections 
equals  0.278,  on  the  average  (calculated  by  us,  using  the 
data  in  table  5  from  Shurkhal  and  others  1991b).  This  is 
almost  two  times  larger  than  that  between  Eurasian  stone 
pines  (D  =  0.138). 

The  lack  of  species-specific  diagnostic  loci  among  stone 
pines  also  confirms  the  correctness  of  uniting  them  in  the 
same  separate  subsection  Cembrae.  Only  Siberian  and 
Korean  stone  pines,  on  locus  Gdh,  and  Swiss  and  Korean 
stone  pines,  on  locus  Mdh-4,  do  not  have  common  alleles. 
Besides,  Korean  stone  pines  have  a  unique  electrophoretic 
pattern  of  LAP  enzyme  indicating  the  existence  (or  expres- 
sion) of  an  additional  Lap-4  locus. 

Loci  Adh-1,  Gdh,  Got-3,  Mdh-4,  and  Pgi-2  make  a  most 
significant  contribution  to  interspecific  differentiation  of 
stone  pines,  but  totally  only  39  percent  of  the  whole  sub- 
section Cembrae  genetic  variation  belongs  to  interspecific 
differences  (table  2). 

Thus,  our  data  give  a  strong  genetic  evidence  of  common, 
apparently  monophyletic  stone  pines  origin  and  compara- 
tively recent  time  of  their  divergence.  Subsection  Cembrae 
seems  to  be  a  valid  taxon. 


24 


Phylogenetic  Relationships  Inferred  From  Genetic 
Distances — Many  publications  have  been  devoted  to  stud- 
ies of  taxonomic  and  phylogenetic  relationships  between 
stone  pines  using  morphological,  physiological,  ecological, 
and  other  traits.  Our  study  was  the  first  one  to  use  genetic 
markers,  isozyme  loci,  for  analysis  of  phylogenetic  relation- 
ships between  all  five  stone  pine  species.  These  species  are 
traditionally  united  in  subsection  Cembrae,  according  to 
some  characteristic  features  (Critchfield  and  Little  1966; 
Farjon  1984;  Little  and  Critchfield  1969;  Shaw  1914).  The 
overwhelming  majority  of  researchers  consider  the  main  spe- 
cific features  of  this  subsection,  indehiscent  cones  and  wing- 
less seeds,  to  be  inherited  fi*om  a  common  ancestor.  These 
traits  seem  to  be  a  relatively  recent  acquisition.  Their  ap- 
pearance is  considered  to  be  a  result  of  coevolution  of  stone 
pines  and  nutcrackers  of  genus  Nucifraga  foraging  their 
seeds  and  playing  the  main  role  in  their  maintenance  and 
dispersal  (Lanner  1980,  1982,  1990;  Tomback  1983; 
Tomback  and  Linhart  1990;  Tomback  and  others  1990). 

Thus,  the  subsection  Cembrae  is  regarded  as  a  natural 
monophyletic  taxon.  Nevertheless,  there  also  exists  a  dif- 
ferent point  of  view  (Critchfield  1986),  according  to  which 
traits  shared  by  all  stone  pines  are  insufficiently  confirmed 
as  homologous.  Data  on  artificial  hybridization  appear  to 
be  rather  contradictory  (Critchfield  1986).  The  subsection 
is  believed  to  have  polyphyletic  origin  and,  therefore,  is 
not  a  valid  taxon  (see  for  discussion,  Lanner  1990).  Millar 
and  Kinloch  (1991)  also  have  pointed  out  a  nimiber  of  im- 
solved  problems  in  taxonomy  and  phylogeny  of  subsection 
Cembrae. 

Molecular  and  biochemical  genetic  markers  have  proven 
to  be  a  useful  tool  for  study  of  biosystematic  and  phyloge- 
netic relationships  in  conifers  (Strauss  and  others  1992). 
We  believe  that  application  of  biochemical  genetic  markers 
like  isozymes  can  help  shed  light  on  the  evolution  of  stone 
pine  species.  The  dendrograms  produced  by  NJM  and 
UPGMA  clustering  using  D^h  and  D  matrix  (table  3)  based 
on  16  isozyme  loci  have  apparently  revealed  real  phyloge- 
netic relationships  between  five  stone  pine  species  (fig.  3). 

According  to  this  dendrogram,  Siberian  and  Swiss  stone 
pines  appear  to  be  a  most  closely  related  pair  of  species 
(D  =  0.065).  This  was  not  imexpected,  because  of  the  well- 
known  surprisingly  high  morphological  similarity  (Bobrov 
1978)  and  successful  artificial  hybridization  (Critchfield 
1986)  between  them.  Practically  all  traditional  botanical 
classifications  note  the  absence  of  diagnostic  morphological 
traits,  which  means  that  interspecific  differences  are  more 
quantitative  than  qualitative.  According  to  paleobotanical 
(mainly  palynological)  data,  Bobrov  (1978)  considered  these 
two  species  as  two  now  separately  located  marginal  parts 
of  a  former  single  ancestral  species  spread  on  a  vast  range 
area,  from  the  western  Alps  to  the  northeastern  Yakutia 
(eastern  Siberia),  whose  pollen  has  been  foimd  in  Paleocene, 
Pleistocene,  and  later  deposits. 

As  Bobrov  (1978)  indicates,  "Siberian  stone  pine  must  be 
regarded  as  the  initial,  more  ancient  species,  formerly  more 
widely  distributed,  whereas  the  West  European  stone  pine, 
Pinus  cembra,  is  evolutionarily  still  very  yoimg  and  has  be- 
come geographically  isolated  in  the  recent  past,  being,  there- 
fore, hardly  disting\iishable  from  its  Siberian  ancestors." 

Breakup  of  this  ancient  range,  its  division  into  two  parts, 
and  decrease  of  its  western  part  occurred,  most  likely,  as 


UPGMA 


Isozyme  loci 


L 


JL 


-L 


225    200    175   150   125    100    75     50  25 


UPGMA 


cpDNA  RFLP 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 

D  (X10-3) 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 


J. 


± 


± 


J. 


90    80     70    60     50    40    30     20  10 


UPGMA 


isozyme  loci 


C3 


d  (X10-5) 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 


_L 


_L 


_L 


I 


405    360  315    270  225    180   135    90  45 


Dch(x10-3) 


NJM  isozyme  loci 

Siberian  stone  pine 


C Siberian 


± 


± 


Swiss  stone  pine 

Whitebari<  pine 

— —  Korean  stone  pine 
-■i-v— i— i^^—  Japanese  stone  pine 

-J  \  \  lD,,(x10-3) 


45    90     135  180    225    270  315  360  405 


NJM 


Swiss  stone  pine 
Siberian  stone  pine 

Korean  stone  pine 


cpDNA  RFLP 

Whitebark  pine 


I  L 


J  L 


15 


30 


45 


65 


Japanese  stone  pine 
-Jd  (X10-5) 


Figure  3 — UPGMA  and  NJM  dendrograms  of  five 
stone  pine  species  (subsection  Cembrae)  based  on 
Cavalli-Sforza  and  Edwards'  (1967)  and  Nei's  (Nei 
1972)  genetic  distances,     and  D,  calculated  from 
isozyme  data,  and  on  d  (Nei  1 987b;  Nei  and  Miller 
1990),  calculated  from  chloroplast  (cp)  DNA  restric- 
tion fragment  length  polymorphism  (RFLP)  data 
(Krutovskii  and  Wagner  1993). 


25 


late  as  in  the  Holocene  and  was  associated  with  its  substi- 
tution in  the  central  part  of  the  area  by  other  species  due 
to  the  global  warming.  The  time  of  divergence  (10,000  to 
25,000  years),  estimated  from  the  genetic  distance  between 
Siberian  and  Swiss  stone  pines  corresponds,  approximately, 
to  these  events  (for  more  details  see  Krutovskii  and  others 
1990).  Thus,  the  low  level  of  divergence  between  Siberian 
and  Swiss  stone  pines  can  be  explained  by  the  short  time 
of  isolation. 

Nevertheless,  a  considerable  amount  of  genetic  difference 
has  accumulated  during  this  time — genetic  distances  be- 
tween Siberian  and  Swiss  stone  pines  are  approximately 
10  times  larger  than  those  between  populations  of  the  same 
species  (table  3)  and  are  comparable  with  distances  reported 
by  Conkle  and  others  (1988)  for  subspecies  and  very  closely 
related  pine  species.  Our  data  have  also  been  recently  con- 
firmed by  Goncharenko  and  others  (1992)  and  Shurkhal  and 
others  (1991a,b;  1992).  Their  estimates  of  Nei's  genetic  dis- 
tance between  Siberian  and  Swiss  stone  pines  equaled 
0.030  and  0.042,  correspondingly.  Thus,  Swiss  stone  pine 
seems  to  be  a  highly  diverged  geographical  race  or  subspe- 
cies of  Siberian  stone  pine,  or  a  species  in  statu  nascendi. 

Unexpectedly  relatively  small  differentiation  found  be- 
tween Japanese  and  Korean  stone  pines  (D  =  0.105)  makes 
us  search  for  a  new  approach  to  understanding  their  origin. 
These  species,  in  fact,  have  little  in  common  in  morphology 
of  both  vegetative  and  generative  organs,  but  according  to 
the  study  of  parenchjnnal  tissue  of  needle  cell  walls,  stone 
pine  species  and  some  closely  related  species  can  be  divided 
into  two  t3T)es:  Siberian  and  Swiss  stone  pine  form  the  first 
type,  while  Japanese  and  Korean  stone  pines  belong  to  the 
second  type,  but  to  different  subtypes  (Litvintseva  1974). 
Perhaps,  the  evolutionary  pathways  of  these  species  have 
truly  separated  from  a  common  ancestor  relatively  recently. 
Another  explanation  of  their  genetic  similarity  is  gene  ex- 
change through  introgressive  hybridization,  which  could 
have  occurred  (or  seldom  occurs),  taking  into  account  the 
present  sympatry  of  these  species  and,  as  a  rule,  poorly 
developed  mechanisms  of  reproduction  isolation  between 
closely  related  pine  species.  More  detailed  discussion  can 
be  foimd  in  Krutovskii  and  others  (1990),  but  recently  ob- 
tained data  also  confirm  the  division  of  Eurasian  stone  pines 
by  isozyme  loci  analysis  into  two  groups:  Siberian  and  Swiss 
stone  pines  (D  =  0.030-0.042),  on  one  hand,  and  Korean  and 
Japanese  (D  =  0.089-0.143),  on  the  other  hand  (Goncharenko 
and  others  1992;  Shurkhal  and  others  1991a,b;  1992). 

A  most  interesting  problem,  from  our  point  of  view,  is  ex- 
planation of  the  whitebark  pine's  closer  affinity  to  Siberian 
and  Swiss  stone  pines  (D  =  0.121-0.130)  than  to  Japanese 
stone  pine  (D  =  0.185),  although  quite  opposite  relation- 
ships could  be  suspected  considering  these  species'  present 
area.  Among  stone  pines  Siberian  and  Swiss  stone  pines 
are,  probably,  genetically  most  similar  to  the  original  an- 
cient stone  pine,  due  to  large  refugiums  in  the  Alps  and 
Siberian  (Altai  and  Sayan)  Mountains,  which  could  help 
them  save  a  great  deal  of  ancient  genetic  variation;  and 
whitebark  pine  also  could  keep  more  genetic  similarity  with 
the  ancient  species  than  Korean  and  Japanese  stone  pines. 
Another  explanation  could  be  that  stabilizing  selection  sup- 
ports genetic  similarity  between  whitebark  pine  and  Siberian 
and  Swiss  stone  pines,  because  whitebark  pine  ecologically 


has,  probably,  more  in  common  with  them  than  with  Korean 
and  Japanese  pines. 

However,  the  dendrogram  obtained  by  NJM,  and  based 
on  Cavalli-Sforza  and  Edwards'  (1967)  distance,  D^^,  has 
revealed  a  kind  of  "intermediate"  position  of  whitebark  pine 
between  two  stone  pine  groups,  Swiss  and  Siberian  stone 
pines,  on  one  hand,  and  Korean  and  Japanese  stone  pines, 
on  the  other  (fig.  3).  In  these  cases  NJM  gives  more  an  in- 
tuitively realistic  phylogenetic  tree.  It  is  important  to  note 
that  both  UPGMA  dendrograms  based  on  D^,,  and  Nei's  dis- 
tance, D,  have  practically  the  same  topology,  but  they  differ 
from  the  NJM  dendrogram.  Thus,  one  can  conclude  that 
assumption  of  equal  evolutionary  rates  of  stone  pine  spe- 
cies can  be  erroneous. 

Millar  and  Kinloch  (1991)  discuss  several  theoretically 
possible  variants  of  whitebark  pine  origin:  (1)  from  ancestral 
Eurasian  stone  pine  with  penetration  into  North  America 
during  the  Mesozoic  (before  Laurasia  segregation).  Ter- 
tiary, or  Quartemary  (through  the  Bering  land  bridge)  time; 
(2)  fix)m  one  of  the  American  pines  of  subsection  Strobi.  Now 
we  can  finally  solve  this  problem.  Nei's  genetic  distances 
(D)  between  whitebark  pine  and  Eurasian  stone  pines  equal 
approximately  0.121-0.256  (table  3).  The  time  of  divergence 
it)  between  whitebark  pine  and  Eurasian  stone  pines,  theo- 
retically estimated  from  these  values,  is  0.6-1.3  million  years, 
using  calibration  coefficient  (k)  5x10^  in  the  equation  t  =  kD 
(Nei  1975;  1987a,b).  This  corresponds  quite  well  to  the  ap- 
proximate geological  time  of  the  Bering  Strait  opening — 
1.8-3.5  million  years  ago — estimated  by  different  methods 
(see  references  in  Grant  1987).  Consequently,  this  event, 
the  Bering  Strait  opening,  can  be,  obviously,  considered 
as  the  beginning  of  genetic  differentiation  between  North 
American  and  Eurasian  pines  due  to  geographical  isolation. 

Thus,  the  hypothesis  of  a  relatively  recent  penetration  of 
ancient  stone  pines  into  North  America  through  the  Bering 
lemd  bridge  in  the  Pliocene  seems  most  preferable.  If  we 
assumed  the  hypothesis  that  a  more  ancient  penetration 
of  stone  pines  through  a  "Euramerican"  connection  had 
occurred  during  Mesozoic  time  (Millar  and  Kinloch  1991), 
much  more  differentiation  between  whitebark  pine  and 
Eurasian  stone  pines  should  have  been  found  in  our  study. 
The  assumption  of  whitebark  pine  origin  from  Strobi  pines 
must  be  rejected,  according  to  our  data,  but  phylogenetic 
relationships  of  stone  pines  with  other  relative  species  from 
subsections  Strobi  and  Cembroides  certainly  need  much 
more  analysis.  The  first  steps  in  this  direction  have  been 
made  recently  using  cpDNA  markers. 

Chloroplast  DNA  Data  and  Pine  Phylogeny— Only 
a  few  studies  on  pine  phylogeny  using  DNA  markers  have 
been  made.  This  allows  us  to  draw  a  short  comparison  of 
the  Millar  and  Kinloch  (1991)  results  with  our  data,  pre- 
sented here,  paying  special  attention  to  phylogenetic  rela- 
tionships of  stone  pines  with  some  other  closely  related 
species. 

Chloroplast  DNA  (cpDNA)  restriction  fragment  length 
polymorphism  (RFLP)  of  the  five  stone  pine  species  and  four 
white  pine  species  of  a  closely  related  subsection  Strobi — 
limber  pine,  sugar  pine  (P.  lambertiana  Dougl.),  western 
white  pine  (P.  monticola  Dougl.),  and  eastern  white  pine — 
have  been  recently  studied  by  Krutovskii  and  Wagner  (1993). 
Data  have  been  obtained  by  molecular  hybridization  of  27 


26 


probes,  representing  the  entire  chloroplast  genome  of 
P.  contorta,  with  restriction  fragments  from  20  endonu- 
cleases.  The  total  nimiber  of  restriction  fragments  per 
sample  varied  from  378  to  422. 

Chloroplast  DNA  variation  among  trees  within  species 
was  not  pronounced  (D  =  0.0-0.0013)  and  was  comparable 
with  variation  between  populations  within  species,  but 
cpDNA  diversity,  in  general,  was  quite  high — 50  to  100 
percent  individual  cpDNA  belonged  to  different  haplotypes 
(for  more  details  see  Krutovskii  and  Wagner  1993).  The 
dendrogram  constructed  on  Nei's  genetic  distances  (D), 
estimated  from  these  data  (Nei  1987b;  Nei  and  Miller 
1990),  has  shown  almost  the  same  phylogenetic  relation- 
ships between  stone  pine  species  as  established  on  isozyme 
loci  data  (fig.  3).  The  stone  and  white  pine  species  have 
been  imdoubtedly  divided  into  two  separate  groups  (fig.  4). 
It  is  interesting  to  note  that  NJM  dendrograms  give  better 
correspondence  between  cpDNA  and  isozyme  data.  Com- 
paring NJM  and  UPGMA  dendrograms  we  can  suppose 
that  assumption  of  homogeneity  of  evolutionary  rates  is 
less  proper  for  cpDNA  than  for  nuclear  (isozyme)  genes. 

The  cpDNA  data  also  indicate  that  genetic  differentia- 
tion is  more  developed  among  the  four  Strobi  species  than 


among  the  Cembrae  species.  The  same  concliision  has  been 
made  by  Shurkhal  and  others  1991a,b;  1992)  on  the  isozyme 
data. 

Chloroplast  DNA's  of  whitebark  pine  and  Eurasian  stone 
pine  species  were  remarkably  similar.  This  similarity,  to- 
gether with  clear  cpDNA  divergence  between  the  two  sub- 
sections, confirms  systematic  classification  of  whitebark 
pine  in  subsection  Cembrae,  and  not  in  subsection  Strobi. 
Phylogenetic  relationships  of  whitebark  pine  with  Siberian 
and  Swiss  stone  pines  revealed  by  cpDNA  RFLP  analysis 
are  surprisingly  consistent  with  the  conclusions  from  the 
analysis  of  16  isozyme  loci  (fig.  3).  Thus,  ovir  observation 
is  not  an  artifact,  because  two  independent  sets  of  genetic 
markers  (isozymes  and  cpDNA  RFLPs)  have  produced  the 
same  conclusions.  Moreover,  Szmidt  (1991)  notes  that  the 
present  geographic  distribution  of  some  forest  tree  species 
often  shows  little  correlation  with  the  degree  of  cpDNA  dif- 
ferentiation among  them,  because  the  present  geographic 
distribution  of  some  species  is  of  relatively  recent  origin. 

The  data  obtained  by  Szmidt  and  others  (1988),  who  used 
four  restriction  enzymes  for  cpDNA  RFLP  analysis  of  12 
pine  species,  including  two  stone  pine  species  (Swiss  stone 
and  whitebark  pines)  and  two  white  pine  species  (western 


UPGMA 


cpDNA  RFLP 


Whitebark  pine 
Swiss  stone  pine 
Siberian  stone  pine 
Korean  stone  pine 
Japanese  stone  pine 
Western  white  pine 
Limber  pine 
Sugar  pine 
Eastern  white  pine 

d  (xlO-3) 


Stone 
Pines 


White 
Pines 


30 


20 


10 


NJM 


J  Sit 

it 


Siberian  stone  pine 
Swiss  stone  pine 
Whitebark  pine 
Korean  stone  pine 
apanese  stone  pine 

Western  white  pine 
Limber  pine 
Sugar  pine 


Stone 
Pines 


Eastern  white  pine 
I  id(x10-3) 


White 
Pines 


10 


20 


30 


Figure  4 — UPGMA  and  NJM  dendrograms  of  five  stone  pine  spe- 
cies (subsection  Cembrae)  and  four  white  pine  species  closely  re- 
lated to  them:  Pious  flexilis,  P.  lambertiana,  P.  monticola,  and  P. 
strobus  (subsection  Strobi),  based  on  Nei's  genetic  distances,  d, 
(Nei  1987b;  Nei  and  Miller  1990),  and  calculated  from  chloroplast 
(cp)  DNA  restriction  fragment  length  polymorphism  (RFLP)  data 
(Krutovskii  and  Wagner  1993). 


27 


and  eastern  white  pines),  also  confirm  our  results.  They 
evaluated  phylogenetic  relationships  between  these  species 
using  UPGMA  clustering  based  on  dissimilarity  index,  anal- 
ogous to  Nei's  D.  According  to  the  dendrogram  given  by 
Szmidt  £ind  others  (1988),  Swiss  stone  and  whitebgirk  pines 
form  a  very  tight  separate  cluster.  Western  white  pine  is 
distantly  linked  to  this  cluster,  which  exactly  corresponds 
to  our  data  (fig.  4).  Eastern  white  pine  forms  a  separate 
branch  but  is  closely  linked  to  the  group  of  Swiss  stone, 
whitebark,  and  western  white  pines. 

Analogous  data  have  been  obtained  by  Strauss  and 
Doerksen  (1990)  who  studied  one  sample  fi"om  each  of  18 
to  19  pine  species  representing  14  to  15  subsections  using 
eight  restriction  enzjnnes  and  17  cloned  DNA  fragments 
(mostly  fi"om  Pseudotsuga  menziesii).  At  least  65  percent 
of  the  chloroplast  genome  was  covered  by  the  combination 
of  the  probes  used.  The  samples  have  also  included  two 
stone  pine  species  (Korean  stone  and  whitebark  pines)  and 
two  white  pine  species  (blue  and  sugar  pines).  Distances 
between  the  samples  were  estimated  via  shared  characters. 
The  stone  £uid  white  pine  species  were  extremely  similar 
to  one  another,  with  distances  ranging  fi^om  0.0  to  0.018. 
They  form  one  cluster.  The  authors  have  mentioned  (data 
have  not  been  presented)  that  other,  more  limited,  restric- 
tion fi'agment  studies  of  section  Strobus,  including  three 
stone  (Swiss,  Siberian,  and  Japanese)  and  three  white  pines 
(Armand's  [P.  armandii  Franch.],  limber,  and  western  white), 
have  given  the  same  indication  of  high  uniformity.  Unfor- 
tunately, Strauss  and  Doerksen  (1990)  have  not  been  able 
to  distinguish  stone  and  white  pines,  probably  because  a 
limited  number  of  restriction  enzymes  or  DNA  probes  have 
been  used  in  their  research. 

Certainly,  stone  pines  and  other  closely  related  species 
need  more  genetic  analysis  using  both  isozyme  and  DNA 
markers.  Many  fascinating  aspects  of  their  evolution  are 
sure  to  remain  hidden,  and  many  discoveries  are  still  w£dt- 
ing  for  research. 

CONCLUSIONS 

The  main  conclusions  based  on  our  data  are: 

1.  Both  biochemical  (isozymes)  and  molecular  (cpDNA) 
genetic  markers  are  very  useful  and  informative  for  study 
of  population  genetic  structure  and  phylogeny  of  stone  pine 
and  other  closely  related  species,  and  they  complement  each 
other. 

2.  In  spite  of  the  high  level  of  isozyme  loci  and  cpDNA 
diversity  among  stone  pine  species,  genetic  differentiation 
between  the  populations  is  not  pronounced,  but  specific. 
As  a  minimum,  the  genetic  differentiation  reflects  their 
geographical  and  ecological  peculiarities.  Thus,  it  is  neces- 
sary to  study  populations  fi'om  the  whole  of  the  stone  pine 
species  area  for  careful  description  of  gene  distribution, 
taking  into  account  even  very  small  genetic  differences 
between  populations  for  developing  gene  conservation  and 
reforestation  programs. 

3.  According  to  both  isozyme  and  cpDNA  restriction  fi'ag- 
ment  analyses,  P.  albicaulis  undoubtedly  belongs  to  Cembrae 
species  and  not  to  Strobi  species. 

4.  Stone  pine  species  consist  of  a  compact  group  of  very 
closely  related  and  genetically  similar  species,  obviously 


having  monophyletic  origin,  supposedly,  fi*om  ancient 
Siberian  stone  pine.  Thus,  subsection  Cembrae  should 
be  considered  as  a  valid  taxon. 

ACKNOWLEDGMENTS 

We  are  very  gratefiil  to  V.  N.  Vorobjev  and  N.  A.  Vorobjeva 
(Institute  of  Ecology  of  Natural  Complexes,  Tomsk,  Russia), 
L.  I.  Milyutin  and  his  laboratory  staff  (Institute  of  Forest 
and  Timber,  Krasnoyarsk,  Russia),  G.  R.  Furnier  (Univer- 
sity of  Minnesota,  Minneapolis,  MN,  U.S.A.),  R,  T.  Gut  (For- 
est Technology  Institute,  Lvov,  Ukraine),  and  P.  Razumov 
(Far  East  Forest  Institute,  Khabarovsk,  Russia)  for  their 
help  in  collecting  and  kindly  providing  Cembrae  species  seed 
samples  for  isozjnne  analysis.  We  are  th£inkful  to  Yu.  S. 
Belokon'  (Institute  of  General  Genetics,  Moscow,  Russia) 
for  very  helpfiil  assistance  in  electrophoretic  ansdysis,  R.  M. 
Lanner  (Utah  State  University,  Logan,  UT,  U.S.A.),  D.  F. 
Tomback  (University  of  Colorado  at  Denver,  Denver,  CO, 
U.S.A.),  and  C.  Millar  (Institute  of  Forest  Grenetics,  Berkeley, 
CA,  U.S.A.)  for  very  useful  comments  on  pine  phylogeny. 
We  also  appreciate  H.-R.  Gregorius'  invaluable  comments 
that  helped  to  substantially  improve  the  manuscript. 

REFERENCES 

Altukhov,  Y.  P.  1990.  Population  genetics:  diversity  and  sta- 
bility. New  York:  Harwood  Academic  Publishers.  352  p. 

Altukhov,  Y.  P.;  Krutovskii,  K  V.;  Dukharev,  V.  A.; 
Larionova,  A.  Y.;  Politov,  D.  V.;  Ryabokon',  S.  M.  1989. 
Biochemicgd  population  genetics  of  woody  forest  species. 
In:  Iroshnikov,  A.  I.,  ed.  International  symposium  on  for- 
est genetics,  breeding  and  physiology  of  woody  plants: 
proceedings;  1989  September  24-30;  Voronezh.  Moscow: 
The  State  Forest  Committee  of  the  USSR:  21-29. 

Bergmginn,  F.  1978.  The  allelic  distribution  of  an  acid  phos- 
phatase locus  in  Norway  spruce  {Picea  abies)  £dong  simi- 
lar climatic  gradients.  Theoretical  and  Applied  Genetics. 
24:  57-64. 

Bobrov,  E.  G.  1978.  Forest-forming  conifers  of  the  USSR. 
Moscow:  Nauka.  190  p.  [In  Russian]. 

Carsey,  K.  S.;  Tomback,  D.  F.  1992.  Growth  form  distribu- 
tion and  genetic  relationships  in  tree  clusters  of  Pinus 
flexilis,  a  bird-dispersed  pine.  Oecologia.  [In  press.] 

Cavalli-Sforza,  L.  L.;  Edwards,  A.  W.  F.  1967.  Phylogenetic 
analysis:  models  and  estimation  procedures.  American 
Journal  of  Human  Genetics.  19(2):  233-257. 

Conkle,  M.  T.;  Schiller,  G.;  Grunwald,  C.  1988.  Electro- 
phoretic analysis  of  diversity  and  phylogeny  of  Pinus 
brutia  £ind  closely  related  taxa.  Systematic  Botany. 
13(3):  411-424. 

Crawford,  D.  J.  1983.  Phylogenetic  and  systematic  infer- 
ences fi'om  electrophoretic  studies.  In:  Tanksley,  S.  D.; 
Orton,  T.  J.,  eds.  Isoz)nnes  in  plant  genetics  and  breed- 
ing. Part  A.  Amsterdam:  Elsevier  Scientific  Publishers: 
257-287. 

Critchfield,  W.  B.  1986.  Hybridization  and  classification 
of  the  white  pines  (Pinus  section  Strobus).  Taxon.  35(4): 
647-656. 

Critchfield,  W.  B.;  Little,  E.  L.  1966.  Geographic  distribu- 
tion of  the  pines  of  the  world.  Washington,  DC:  U.S.  De- 
partment of  Agriculture.  99  p. 


28 


Dancik,  B.  P.;  Yeh,  F.  C.  1983.  Allozyme  variability  and 
evolution  of  lodgepole  pine  {Pinus  contorta  var.  latifolia) 
and  jack  pine  (Pinus  banksiana)  in  Alberta.  Canadian 
Journal  of  Genetics  and  Cytology.  25:  57-64. 

Farjon,  A.  1984.  Pines:  drawings  and  descriptions  of  the 
genus  Pinus.  Leiden:  E.  J.  Brill.  204  p. 

Farris,  J.  S.  1972.  Estimating  phylogenetic  trees  from  dis- 
tance matrices.  American  Naturalist.  106(951):  645-668. 

Fumier,  G.  R.;  Adams,  W.  T.  1986.  Geographic  patterns  of 
allozyme  variation  in  Jeffrey  pine.  American  Journal  of 
Botany.  73(7):  1009-1015. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P.  1987. 
Effect  of  avian  seed  dispersal  on  the  genetic  structure  of 
whitebark  pine  populations.  Evolution.  41(3):  607-612. 

Goncharenko,  G.  G.;  Padutov,  V.  E.;  Silin,  A.  E.  1992.  Popu- 
lation structure,  gene  diversity,  and  differentiation  in  natu- 
ral populations  of  cedar  pines  (Pinus  subsection  Cembrae, 
Pinaceae)  in  the  USSR.  Plant  Systematics  and  Evolu- 
tion. 182(3-4):  121-134. 

(rottlieb,  L.  D.  1977.  Electrophoretic  evidence  and  plant 
systematics.  Annals  of  Missouri  Botanical  Garden.  64: 
161-180. 

Grant,  W.  S.  1987.  Genetic  divergence  between  congeneric 
Atlantic  and  Pacific  Ocean  fishes.  In:  Ryman,  N.;  Utter, 
F.,  eds.  Population  genetics  and  fishery  management. 
Seattle,  WA:  University  of  Washington  Press:  225-246. 

Hamrick,  J.  L.  1983.  The  distribution  of  genetic  variation 
within  and  among  natural  plant  populations,  hi:  Chambers, 
S.;  Schonewald-Cox,  C,  eds.  Genetics  and  wild  population 
management.  Reading,  PA:  Addison- Wesley:  335-349; 
501-509. 

Hamrick,  J.  L.;  Godt,  M.  J.  1989.  Allozyme  diversity  in  plant 
species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler,  A.  L.; 
Weir,  B.  S.,  eds.  Plant  population  genetics  breeding  and 
genetic  resources.  Sanderland,  MA:  Sinauer  Associates: 
43-63. 

Hamrick,  J.  L.;  Loveless,  M.  D.  1986.  The  influence  of  seed 
dispersal  mechanisms  on  the  genetic  structure  of  plant 
populations.  In:  Estrada,  J.;  Fleming,  T.  H.,  eds.  Frugi- 
vores  and  seed  dispersal.  Dordrecht,  The  Netherlands: 
Dr.  W.  Junk  Publishers:  211-223. 

Hamrick,  J.  L.;  Mitton,  J.  B.;  Linhart,  G.  B.  1981.  Levels  of 
genetic  variation  in  trees:  influence  of  life  history  charac- 
teristics. In:  Conkle,  M.  T.,  ed.  Symposium  on  isozymes 
in  North  American  forest  trees  and  forest  insects:  proceed- 
ings. Berkeley,  CA:  U.S.  Department  of  Agricultiu-e,  For- 
est Service:  35-41. 

Krutovskii,  K.  V.;  PoHtov,  D.  V.;  Altukhov,  Yu.  P.  1987. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  I.  Genetic  control  of  isozyme  systems.  Genetika. 
23(12):  2216-2228.  [In  Russian].  [English  translation  pub- 
lished by  Plenvun  Publishing  Corp.  in  Soviet  Genetics; 
1988  June:  1568-1576]. 

Krutovskii,  K.  V.;  Pohtov,  D.  V.;  Altukhov,  Yu.  P.  1988. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  II.  Level  of  allozyme  variability  in  western 
Sayan.  Genetika.  24(1):  118-124.  [In  Russian].  [English 
translation  published  by  Plenum  Publishing  Corp.  in 
Soviet  Genetics;  1988  July:  81-86]. 

Krutovskii,  K  V.;  Pohtov,  D.  V.;  Altukhov,  Yu.  P.;  Milyutin, 
L.  I.;  Kuznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorobjev,  V.  N.; 
Vorobjeva,  N.  A.  1989.  Genetic  variability  in  Siberian 
stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic  diversity 


and  differentiation  between  poptilations.  Genetika.  25(11): 
2009-2032.  [In  Russian].  [EngHsh  translation  published 
by  Plenum  Publishing  Corp.  in  Soviet  Genetics;  1990 
May:  1343-1362]. 

Krutovskii,  K.  V.;  PoHtov,  D.  V.;  Altukhov,  Yu.  P.  1990.  In- 
terspecific genetic  differentiation  of  Eurasian  stone  pines 
for  isoenzyme  loci.  Genetika.  26(4):  694-707.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1990  October:  440-450]. 

Krutovskii,  K.  V.;  Politov,  D.  V.  1992.  Study  of  intra-  and 
interspecific  genetic  differentiation  between  E\u*asian 
cedar  pines  using  isozyme  loci  and  multidimensional 
analysis.  In:  Symposium  on  molecular  mechanisms  of 
genetic  processes:  proceedings.  Moscow:  Nauka  Publish- 
ers: 87-96.  [In  Russian]. 

Krutovskii,  K.  V.;  Wagner,  D.  B.  1993.  [Unpubhshed.]  Chlo- 
roplast  DNA  variation  and  phylogeny  in  Pinus  subsec- 
tions Cembrae  and  Strobi  (subgenus  Strobus).  Lexington, 
KY:  University  of  Kentucky,  Department  of  Forestry. 
[Working  draft]. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in  the 
ecology  and  evolution  of  limber  and  whitebark  pines.  In: 
6th  North  American  Forest  Biology  Workshop:  proceed- 
ings; 1980  August  11-13;  Edmonton,  AB.  Edmonton,  AB: 
University  of  Alberta:  15-47. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for  seed 
dispersal  by  Clark's  nutcracker.  Canadian  Journal  of 
Forest  Research.  12:  391-402. 

Lanner,  R.  M.  1990.  Biology,  taxonomy,  evolution,  and  geog- 
raphy of  stone  pines  of  the  world.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — sjonposium 
on  whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Litvintseva,  M.  V.  1974.  Characteristics  of  needle  paren- 
chymal cell  structure  in  species  of  the  group  Cembrae 
of  the  genus  Pinus.  Botanicheski  Zhurnal.  [Leningrad]. 
59(10):  1501-1505.  [In  Russian]. 

Little,  E.  L.;  Critchfield,  W.  B.  1969.  Subdivisions  of  the 
genus  Pinas  (pines).  Res.  Rep.  1144.  Washington,  DC: 
U.S.  Department  of  Agriculture,  Forest  Service:  1-51. 

Loveless,  M.  D.;  Hamrick,  J.  L.  1984.  Ecological  determi- 
nants of  genetic  structure  in  plant  populations.  Annual 
Review  of  Ecology  and  Systematics.  15:  65-95. 

Millar,  C.  I.  1983.  A  steep  cline  in  Pinus  muricata.  Evolu- 
tion. 32(2):  311-319. 

Millar,  C.  I.;  Kinloch,  B.  B.  1991.  Taxonomy,  phylogeny,  and 
coevolution  of  pines  and  their  stem  rusts.  In:  Hiratsuka,  Y.; 
Samoil,  J.  K.;  Blenis,  P.  V.;  Crane,  P.  E.;  Laishley,  B.  L., 
eds.  Rusts  of  pine.  Inf.  Rep.  NOR-X-317.  Edmonton,  AB, 
Canada:  Northern  Forestry  Centre. 

Nei,  M.  1972.  Genetic  distance  between  populations. 
American  Naturahst.  106:  283-292. 

Nei,  M.  1975.  Molecular  population  genetics  and  evolution. 
New  York:  American  Elsevier.  288  p. 

Nei,  M.  1977.  F-statistics  and  analysis  of  gene  diversity  in 
subdivided  populations.  Annals  of  Human  Genetics.  41: 
225-233. 

Nei,  M.  1978.  Estimation  of  average  heterozygosity  and 
genetic  distance  from  a  small  niunber  of  individuals. 
Genetics.  89:  583-590. 


29 


Nei,  M.  1987a.  Genetic  distance  and  molecular  phylogeny.  In: 
Ryman,  N.;  Utter,  F.,  eds.  Population  genetics  and  fishery 
management.  Seattle,  WA:  University  of  Washington 
Press:  193-223. 

Nei,  M.  1987b.  Molecular  evolutionary  genetics.  New  York: 
Columbia  University  Press.  512  p. 

Nei,  M.;  Miller,  J.  C.  1990.  A  simple  method  for  estimating 
average  nvunber  of  nucleotide  substitutions  within  and 
between  populations  from  restriction  data.  Genetics.  125: 
873-879. 

Politov,  D.  V.  1989.  Allozyme  polymorphism,  genetic  differ- 
entiation, and  mating  system  in  Siberian  stone  pine,  Pinus 
sibirica  Du  Tour.  Moscow:  Institute  of  General  Genetics, 
Russian  Academy  of  Sciences.  190  p.  Dissertation  Thesis. 
[In  Russian]. 

Politov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1989. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  III.  Linkage  relationships  among  isozyme  loci. 
Genetika.  25(9):  1606-1618.  [In  Russian].  [Enghsh  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Soviet 
Genetics;  1990  March:  1053-1063]. 

Politov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variability 
in  Siberian  stone  pine,  Pinus  sibirica  Du  Tour.  V.  Analy- 
sis of  mating  system.  Genetika.  26(11):  2002-2011.  [In 
Russian].  [English  translation  published  by  Plenum  Pub- 
Hshing  Corp.  in  Soviet  Genetics;  1991  May:  1309-1316]. 

Politov,  D.  v.;  Krutovskii,  K  V.;  Altukhov,  Yu.  P.  1992.  Iso- 
zyme loci  characteristics  of  gene  banks  of  populations 
of  cedar  pines.  Genetika.  28(1):  93-114.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1992  July:  76-95]. 

Rogers,  J.  S.  1972.  Measures  of  genetic  similarity  and  ge- 
netic distance.  In:  Studies  in  Genetics  VII.  Publ.  7213. 
Austin,  TX:  University  of  Texas:  145-153. 

Rohlf,  F.  J.  1988.  NTSYS-pc:  numerical  taxonomy  and  mul- 
tivariate analysis  system.  New  York:  Exeter  Publishing 
Ltd. 

Saitou,  N.;  Nei,  M.  1987.  The  neighbor-joining  method:  a 
new  method  for  reconstructing  phylogenetic  trees.  Molecu- 
lar Biology  and  Evolution.  4:  406-425. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1991.  Relatedness  within 
clusters  of  bird-dispersed  pine  and  the  potential  for  kin 
interactions.  Heredity.  67:  41-48. 

Shaw,  G.  R.  1914.  The  genus  Pinus.  Publ.  Arnold  Arboretimi 
No.  5.  Cambridge,  IVLA:  Riverside  Press.  96  p. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1991a.  Genetic 
differentiation  of  18  pine  species  on  allozyme  loci,  genus 
Pinus,  subgenus  Strobus  and  Pinus.  Doklady  Akademii 
Nauk  SSSR.  316(2):  484-488.  [In  Russian].  [English  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Doklady 
Biological  Sciences;  1991]. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1991b.  Phyloge- 
netic analysis  of  genus  Pinus  by  allozyme  loci:  genetical 
differentiation  of  subgenus.  Genetika.  27(5):  1193-1205. 
[In  Russian].  [English  translation  published  by  Plenum 
Publishing  Corp.  in  Soviet  Genetics;  1991  November]. 

Shurkhal,  A.;  Podogas,  A.;  Zhivotovsky,  L.  1992.  Allozyme 
differentiation  in  the  genus  Pinus.  Silvae  Genetica.  41(2): 
105-109. 

Sneath,  P.  H.  A.;  Sokal,  R.  R.  1973.  Numerical  taxonomy. 
San  Francisco:  W.  H.  Freeman:  230-234. 


Steinhoff,  R.  J.;  Joyce,  D.  G.;  Fins,  L.  1983.  Isozyme  varia- 
tion in  Pinus  monticola.  Canadian  Journal  of  Forest  Re- 
search. 13(6):  1122-1131. 

Strauss,  S.  H.;  Doerksen,  A.  H.  1990.  Restriction  fragment 
analysis  of  pine  phylogeny.  Evolution.  44(4):  1081-1096. 

Strauss,  S.  H.;  Bosquet,  J.;  Hipkins,  V.  D.;  Hong,  Y.-P. 
1992.  Biochemical  and  molecular  genetic  markers  in  bio- 
systematic  studies  of  forest  trees.  New  Forests.  6: 125-158. 

Swofford,  D.  L.;  Selander,  R.  B.  1981.  BIOSYS-1:  a  FOR- 
TRAN program  for  the  comprehensive  analysis  of  elec- 
trophoretic  data  in  population  genetics  and  systematics. 
Journal  of  Heredity.  72:  281-283. 

Szmidt,  A.  E.  1991.  Phylogenetic  and  applied  studies  on 
chloroplast  genome  in  forest  conifers.  In:  Hattemer,  H.  H.; 
Fineschi,  S.;  Cannata,  F.;  Malvolti,  E.  V.,  eds.  Biochemi- 
cal markers  in  the  population  genetics  of  forest  trees. 
The  Hague,  The  Netherlands:  SPB  Academic  Publishing: 
185-196. 

Szmidt,  A.  E.;  Sigurgeirsson,  A.;  Wang,  X.  R.;  Hdllgren, 
J.  E.;  Lindgren,  D.  1988.  Genetic  relationships  among 
Pinus  species  based  on  chloroplast  DNA  pol3miorphism. 
In:  H^lgren,  J.-E.,  ed.  Frans  Kempe  Sympositun  in  Umeli 
on  molecular  genetics  of  forest  trees:  proceedings;  1988; 
Umed,  Sweden.  Ume&,  Sweden:  Swedish  University  of 
Agricultural  Sciences:  33-47. 

Tomback,  D.  F.  1983.  Nutcrackers  and  pines:  coevolution  or 
coadaptation?  In:  Nitecki,  N.  H.,  ed.  Coevolution.  Chicago: 
University  of  Chicago  Press:  179-223. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution of  whitebark  pine  and  nutcrackers:  implications  for 
forest  regeneration.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposiimi  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
movmtain  Research  Station:  118-129. 

Tomback,  D.  F.;  Holtmeier,  F.  K.;  Mattes,  H.;  Carsey,  K.  S.; 
Powell,  M.  L.  1992.  Tree  clusters  and  growth  form  distri- 
bution in  Pinus  cembra,  a  bird-dispersed  pine.  Denver, 
CO:  University  of  Colorado  at  Denver,  Department  of 
Biology.  [Working  draft]. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Wheeler,  N.  C;  Guries,  R.  P.  1982.  Population  structure, 
genetic  diversity  and  morphological  variation  in  Pinus 
contorta  Dougl.  Canadian  Journal  of  Forest  Research. 
12(3):  595-606. 

Wheeler,  N.  C;  Guries,  R.  P.  1987.  A  quantitative  measure 
of  introgression  between  lodgepole  and  jack  pines.  Cana- 
dian Journal  of  Botany.  65:  1876-1885. 

Wheeler,  N.  C;  Guries,  R.  P.;  O'Malley,  D.  M.  1983.  Bio- 
systematics  of  the  genus  Pinus,  subsection  Contortae. 
Biochemical  Systematics  and  Ecology.  11:  333-340. 

Yeh,  F.  C.  H.;  Cheliak,  W.  M.;  Dancik,  B.  P.;  Illingworth,  K; 
Trust,  D.  C;  Pryhitka,  B.  A.  1985.  Population  differentia- 
tion in  lodgepole  pine,  Pinus  contorta  ssp.  latifolia:  a  dis- 
criminant analysis  of  allozyme  variation.  Canadian  Jour- 
nal of  Genetics  and  Cytology.  27(2):  210-218. 


30 


COEVOLUTIONAL  ASPECTS  OF  STONE 
PINES  AND  NUTCRACKERS 

Hermann  Mattes 


Abstract — ^Nutcrackers  (Nucifraga  caryocatactes  and  N. 
Columbiana)  have  developed  a  strong  mutualistic  relationship  to 
stone  pines.  Seeds  of  stone  pines  are  the  most  important  food  for 
at  least  9  months  of  the  year  and  for  raising  the  young.  In  addi- 
tion to  special  adaptations  on  gathering,  transporting,  caching, 
and  finding  again  the  hoarded  seeds,  the  whole  annual  cycle  of 
the  nutcracker's  life  (time  of  breeding  and  moulting),  its  mating 
system,  and  its  habitat  use  are  adjusted  to  the  use  of  pine  seeds. 


Within  the  family  of  jays,  magpies,  and  crows  (Cor- 
vidae)  many  characteristics  of  preadaptive  value  for  seed 
dispersal  occur  (Tomback  and  Linhart  1990;  Turcek  and 
Kelso  1968).  Three  corvid  species  have  evolved  a  strong 
mutualistic  system  with  pines:  Clark's  nutcracker  (Nuci- 
fraga Columbiana)  and  the  pinyon  jay  (Gymnorhinus 
cyanocephalus)  in  North  America  and  the  Exiropean  nut- 
cracker {Nucifraga  caryocatactes)  in  Eurasia.  Compre- 
hensive studies  on  food  regime  and  seed  dispersal  were 
carried  out  by  many  authors  (Bibikov  1948;  Crocq  1990; 
Hayashida  1982,  these  proceedings;  Marzlviif  and  Balda 
1992;  Mattes  1978;  Reijmers  1959;  Saito  1983;  Swanberg 
1951;  Tomback  1977;  Vander  Wall  and  Balda  1977,  1981). 
Coevolution  in  corvids  and  pines  does  not  only  refer  to 
feeding  and  caching  behavior,  but  the  whole  life  cycle  is 
adjusted  to  the  use  of  the  special  food  resource.  This  pa- 
per describes  the  process  for  the  European  nutcracker. 

THE  EUROPEAN  NUTCRACKER 

The  annual  life  cycle  (fig.  1)  of  the  nutcracker  is  best 
looked  at  fi'om  August  onward.  Unripe  seeds  of  the  Swiss 
stone  pine  (cembra  pine,  Pinus  cemhra)  are  eaten  by  the 
nutcracker  throughout  August  or  even  in  July;  thus,  the 
bird  will  not  miss  the  point  when  the  seeds  are  fully  de- 
veloped. At  this  time  the  resin  flow  in  the  cone  stops,  and 
cone  scales  get  loose.  Now  the  seeds  can  be  easily  taken 
out  of  the  cone,  and  the  nutcracker  begins  gathering  and 
caching  seeds  for  winter  supply.  Pine  seed  gathering  will 
last  until  all  seeds  have  been  harvested  and  almost  no 
cones  are  left  on  the  trees.  Only  twice  within  19  years  of 
investigation  in  the  Engadine  a  cone  crop  lasted  until  the 
beginning  of  the  following  spring. 

In  the  Alps,  nutcrackers  carry  seeds  of  cembra  pine  over 
distances  up  to  15  km  (including  differences  in  altitude  of 
at  least  700  m)  back  to  their  territories.  They  carry  45 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Hermann  Mattes  is  Professor  of  Landscape  Ecology  and  Geography, 
Department  of  Geography,  University  of  Miinster,  Germany. 


pine  seeds  per  flight  on  an  average;  however,  a  maximum 
load  of  134  cembra  pine  seeds  has  been  found  in  the  sub- 
lingual pouch  of  a  nutcracker.  Main  morphological  and 
ethological  adaptations  for  seed  use  are: 

•  a  long  and  pointed  bill  (fig.  2)  that  enables  the  bird 

to  take  seeds  out  of  the  cones  as  one  would  with  twee- 
zers (most  other  corvids  have  a  hooked  bill); 

•  a  rist  in  the  lower  mandible  allows  the  bird  to  fix  and 
crack  pine  seeds,  supported  by  a  short-cut  tongue; 

•  the  subhngual  pouch  allows  carrying  a  large  quantity 
of  seeds  (fig.  3)  without  disturbing  other  functions  of 
the  bill; 

•  ability  to  test  seed  quality  by  means  of  seed  color  and 
resonance  (by  so-called  bill-clicking). 

SEED  CACHES 

The  cached  seeds  provide  the  main  food  source  for  the 
nutcracker  from  November  to  April,  and  to  a  lesser  degree 
until  July.  A  single  nutcracker  stores  more  than  25,000 
seeds.  As  yet  it  has  not  been  clearly  shown  how  the  nut- 
cracker can  retrieve  its  seed  caches.  However,  as  can  be 
concluded  fi'om  observations  in  the  field  and  in  the  labora- 
tory, nutcrackers  apparently  remember  each  single  cache 
(Tomback  1982,  and  others).  Further,  there  are  no  spe- 
cific features  of  caches,  although  nutcrackers  obviously 
prefer  to  deposit  seeds  near  border  lines  and  any  changes 
of  surface  structure.  Experiments  with  the  Clark's  nut- 
cracker by  Kamil  and  Balda  (1988)  and  Vander  Wall 
(1982)  showed  that  nutcrackers  orient  (navigate)  opti- 
cally, as  expected,  and  additionally  with  a  magnetic  sense. 
Understanding  cache  findings,  however,  requires  more  de- 
tailed information,  as  nutcrackers  detect  caches  even  un- 
der a  thick  cover  of  snow.  Balda  and  Kamil  (1989)  sug- 
gest an  overall  fuzzy  optical  navigation  for  finding  caches. 
This  wovild  allow  the  birds  to  locate  a  cache  even  if  there 
were  distinctive  changes,  for  example,  in  height  of  snow 
cover,  broken  branches,  or  fallen  trees.  It  is  still  an  open 
question  how  the  bird  can  find  caches  under  a  snow  cover 
of  20  cm  and  higher,  getting  very  directly  to  the  hoarded 
seeds.  Snow  tunnels  of  130  cm  (Burckhardt  1958)  and 
about  300  cm  (R.  Stem,  personal  communication)  were 
reported,  and — what  is  most  surprising — the  nutcrackers 
did  find  the  cached  seeds.  Crocq  (1990)  found  the  nut- 
crackers leaving  their  subalpine  habitat  when  snow  cover 
reached  170  cm  on  imdistixrbed  surfaces. 

It  shoiild  be  emphasized  that  nutcrackers  dig  success- 
fully for  their  caches  throughout  winter  imtil  early  spring. 
That  guarantees  sufficient  food  for  the  brood  during  that 
difficult  time  of  the  year.  Nutcrackers  must  remember 
not  only  the  sites  of  the  caches  but  additionally  those  that 
already  have  been  used.  This  doubles  the  efforts  of 


31 


> 

twmm 


11 


Jan.     Feb.    Mar.      Apr.    May    June     July     Aug.    Sept.     Oct.    Nov.  Dec. 

Month 

Figure  1 — Annual  activities  of  the  nutcracker  in  the  Alps  for  years  with  a  medium  cone  crop.  1  -3,  feeding: 
1 ,  seed  harvest  and  transport;  2,  use  of  cached  seeds;  3,  consumption  of  animal  food,  for  example,  spi- 
ders, insects;  4,  winter  period;  5-7  breeding  period:  5,  courtship  and  nest  building;  6,  time  with  clutch  and 
nestlings;  7,  family  groups  with  fledged  young;  8-9,  moulting:  8,  moult  of  adults;  9,  moult  of  fledglings; 
10-11,  movements:  10,  dismigration;  11,  migration. 


memory.  There  is  some  evidence  that  nutcrackers  use 
caches  from  the  penultimate  seed  crop,  which  may  be  im- 
portant in  case  a  cone  crop  fails.  In  years  of  small  seed 
crops,  nutcrackers  often  recache  the  seeds.  Temporary 
caches  are  made  in  cases  of  strong  competition  for  seeds 
(when  great  nimabers  of  nutcrackers  and  squirrels  concen- 
trate on  a  few  cone-bearing  trees).  In  temporary  caches, 
nutcrackers  mostly  hide  cones. 

BREEDING  AND  HABITAT 

Nutcrackers  start  breeding  in  March  (fig.  4).  At  this 
time  mean  daily  temperature  is  under  0  °C.  Early  breed- 
ing is  necessary  because  the  next  generation  has  to  be 
ready  for  harvesting  seeds  in  August  to  ensure  the  winter 
food  supply.  Breeding  takes  about  18  days,  the  nestlings 
fledge  when  24  to  28  days  old,  and  there  is  an  extraordi- 
narily long  time  of  parental  care  for  the  fledglings — until 
they  are  about  100  days  old  (Oberauer  1991;  Rudat  1978). 
Extensive  care  by  the  parents  is  needed  because  of  scarcity 


Figure  2 — Inside  the  nutcracker's  bill:  short 
split  tongue,  underneath  opening  of  sublin- 
gual pouch  with  peanut  in  the  entrance  part, 
rist  in  the  lower  mandible. 


32 


Figure  3 — Nutcracker  with  filled  sublingual 
pouch.  Morteratsch,  Engadin;  September 
1975. 


of  food  up  to  mid-June,  and  to  provide  training  for  caching 
and  digging  for  seeds. 

Courtship  behavior  starts  in  the  end  of  February  or 
early  March.  Synchronizing  with  the  partner  is  of  ex- 
treme importance  for  successful  breeding.  By  February 
the  nutcrackers'  so-called  ceremonial  gatherings  begin 
and  then  can  be  frequently  observed  (Swanberg  1956). 
Up  to  12  or  more  nutcrackers  from  neighboring  and  some- 
times far  territories  meet  in  conspicuous  noisy  groups.  An 
individually  marked  female  nutcracker  that  had  spent  the 
winter  outside  its  territory  came  back  late — March  17.  I 
discovered  her  fighting  for  the  territory  while  a  ceremo- 
nial gathering  was  going  on.  She  succeeded  in  reoccupy- 
ing  her  former  place,  and  raised  two  nestlings  after  hav- 
ing a  three-egg  clutch  with  one  unfertile  and  half-sized 
egg.  Neighbors  know  each  other  individually.  So,  except 
in  artifically  created  feeding  places,  fights  are  very  rare. 

Nutcracker  territories  are  organized  in  a  different  way 
than  those  of  other  passerines.  Only  the  area  very  close 
to  the  nest  is  defended  against  intruders  and  thus  can  be 
called  "territory"  in  the  usual  sense.  In  a  broader  sense, 
the  territory  comprises  the  area  of  food  caches  and  is 
about  5  to  12  ha  large  according  to  habitat.  These  "terri- 
tories" can  overlap  and  are  not  defended  because  cached 
seeds  are  only  accessible  to  the  very  same  nutcracker.  In 
addition,  all  places  regularly  visited  for  seed  gathering 
and  transport  should  be  regarded  as  home  range.  Habi- 
tats of  the  nutcracker  are  characterized  by  evergreen  co- 
nifers. In  such  forests  sheltered  nesting  places  are  avail- 
able, and  because  snow  cover  is  low  under  dense  tree 
canopies  cached  seeds  can  be  easily  recovered. 


Nutcrackers  are  adapted  in  many  ways  for  breeding 
under  winter  conditions.  The  nest  is  heavily  built  with  a 
layer  of  dry  decayed  wood  for  insulation  against  low  tem- 
perature. Beginning  with  the  first  egg,  the  parents  must 
warm  the  clutch  without  long  interruptions.  Breeding 
starts  after  the  last  egg  has  been  laid,  and  egg  tempera- 
ture is  raised  from  about  25  °C  to  breeding  conditions  of 
about  38  °C.  Different  from  most  corvids,  both  adults 
breed.  The  male  develops  a  large  breeding  patch,  too. 
The  female's  breeding  time  is  somewhat  higher  during  the 
day,  and  it  also  stays  on  the  nest  at  night.  The  nestlings 
are  fed  with  pine  seeds  from  their  first  day  of  life.  Addi- 
tional food  consists  of  spiders,  and,  if  available,  insects. 
Nutrient  analysis  in  seeds  of  whitebark  pine  {Pinus 
albicaulis)  by  Lanner  and  Gilbert  (these  proceedings) 
showed  a  high  nutritive  quality  in  caloric  value,  mineral 
nutrients,  fatty  acids,  and  amino  acids.  Should  this  apply 
also  to  cembra  pine  seeds,  it  is  easy  to  understand  why 
first-day  nestlings  can  live  mainly  on  a  seed  diet. 

For  successful  overwintering  and  reproduction,  suffi- 
cient numbers  of  seed  caches  must  be  available  to  both 
parents.  First-year  birds  dismigrate  in  August  and  estab- 
lish their  own  territory  when  seed  caching  starts.  First- 
year  birds  that  are  not  fully  developed  until  August  will 
have  a  minor  chance  to  survive  winter.  A  few  first-year 
birds  leave  their  native  area  in  September  and  migrate  in 
southern  directions  (Mattes  and  Jenni  1984).  Adult  nut- 
crackers are  highly  sedentary.  Nonbreeding  individuals 
might  have  action  ranges  within  or  between  the  territo- 
ries of  breeding  birds.  This  popiilation  reserve  can  com- 
pensate for  winter  losses  of  adult  birds. 

In  the  "Stazer  Wald"  I  observed  a  third  bird  helping  a 
pair  build  their  nest;  unfortunately,  no  further  observa- 
tion was  possible.  Similar  to  several  jay  species,  helping 
in  breeding  may  be  an  evolutional  feature  in  the  nut- 
cracker (Woolfenden  and  Fitzpatrick  1984).  Nutcrackers 
live  monogamously  for  life,  as  can  be  concluded  from  indi- 
vidually marked  birds  (Mattes  1978;  Swanberg  1956).  In 


Figure  4 — A  nutcracker  breeding,  the  nest 
completely  covered  by  snow.  March  1 9, 
1975,  Engadin. 


33 


the  Engadine,  the  mortality  rate  in  adult  birds  was  84  per- 
cent (Mattes,  in  preparation)  on  an  average.  The  oldest 
birds  found  were  at  least  15.  Keeping  territories  and  part- 
ners for  life  supports  synchronization  of  the  parents,  and 
individual  knowledge  of  neighbors  minimizes  social  com- 
petition. Social  hierarchy  is  evident  and  most  conspicu- 
ous at  artifical  feeding  places.  Females  take  over  the  so- 
cial rank  of  their  mates  (Grabher-Meyer  1991). 

In  passerines,  moult  normally  takes  place  after  the 
breeding  season.  Prolonged  care  of  the  fledglings  and 
beginning  of  seed  hoarding  in  August  would  not  leave 
time  for  movdting  in  late  summer.  Actually,  moult  starts 
in  March  with  change  of  remiges  and  tail  feathers,  and 
ends  in  July  when  body  feathers  are  changed.  Moult  com- 
pletely parallels  breeding  season;  the  double  physiological 
effort  causes  a  very  long  moulting  period.  Body  feathers 
are  moulted  relatively  quickly  during  the  warmest  part  of 
the  year. 

REACTION  TO  CHANGES 

Predation  on  pine  seeds  has  influenced  the  whole  life 
cycle  of  the  nutcracker.  Many  features  common  to  the 
corvids  are  particularly  developed  in  the  nutcracker.  Still 
there  is  a  broad  and  flexible  reaction  to  environmental 
changes.  This  is  most  apparent  during  the  invasions  of 
Siberian  nutcrackers  (N.  c.  macrorhynchos)  into  western 
Europe.  Diet  and  feeding  as  well  as  habitat  use  vary 
widely.  Nevertheless,  most  of  the  invading  birds  very 
likely  die.  Some,  however,  return  to  their  breeding  ground 
either  in  the  current  autumn  or  in  the  next  spring.  Nimaer- 
ous  birds  ringed  in  the  Aland  archipelago  (Finland)  in 
August  1968  were  recaptured  a  few  weeks  later  in  the 
East;  three  birds  were  found  beyond  the  Urals  at  a  maxi- 
mum distance  of  almost  3,000  km  (Zink  1981).  Invaders 
stayed  at  wintering  places  in  Bielefeld,  Germany,  when 
Siberian  pine  seeds  (Pinus  sibirica)  were  offered  ad 
libidimi  (Conrads  and  Balda  1979). 

Outside  the  natural  range  of  cembra  pine  (Central  Alps, 
Tatra,  and  a  few  locations  in  the  Carpathian  mountains) 
the  European  nutcracker  normally  feeds  on  hazel  nuts 
iCorylus  avellana).  Slight,  but  statistically  significant, 
differences  in  bill  size  are  apparent  in  central-alpine 
populations.  Nutcrackers  from  the  Engadine  and  migrat- 
ing birds  at  Col  de  Bretolet  in  southwestern  Switzerland 
show  a  narrower  lower  mandible  and  a  longer  bill  com- 
pared with  populations  from  southern  Germany  and  from 
northern  Europe  (Mattes  1978;  Mattes  and  Jenni  1984). 
Also,  there  are  distinct  differences  in  bill  size  in  popula- 
tions of  A^.  c.  macella  from  Szetchuan  and  from  Sikkim 
(Mattes,  unpublished).  Generally,  there  seems  to  be  a 
correlation  of  width  of  lower  mandible  and  size  of  main 
food.  The  distance  between  the  branches  of  the  lower 
mandible  limits  the  size  of  the  entrance  to  the  sublingual 
pouch.  As  the  main  food  of  most  Asian  populations  of  the 
nutcracker  is  not  yet  sufficiently  known,  bill  size  and  bill 
structure  cannot  yet  be  fully  explained.  Furthermore,  bill 
size  in  some  subspecies  may  reflect  feeding  conditions  at 
former  times. 


REFERENCES 

Balda,  Russel  P.;  Kamil,  Alan  C.  1989.  A  comparative 

study  of  cache  recovery  by  three  corvid  species.  Animal 

Behavior.  38:  486-495. 
Bibikov,  D.  I.  1948.  On  the  ecology  of  the  nutcracker. 

Trudy  Pechoro-Ilychkoge  Gosud.  Zapovednik.  4:  89-112. 
Biu-ckhardt,  D.  1958.  Vom  Vorratanlegen  des  Tannen- 

hahers.  Biindnerwald.  11:102-114. 
Conrads,  Klaus;  Balda,  Russell  P.  1979.  Uberwinter- 

imgschancen  Sibirischer  Tannenhaher  {Nucifraga 

caryocatactes  macrorhynchos)  im  Invasionsgebiet. 

Bericht  des  Naturwissenschaftlichen  Vereins  Bielefeld. 

24:  115-137. 

Crocq,  Claude.  1990.  Le  casse-noix  mouchete  {Nucifraga 
caryocatactes).  Paris:  Lechevalier/R.  Chabaud.  326  p. 

Grabher-Meyer,  Arno.  1991.  Beobachtungen  zur  Brut- 
biologie  einer  kleinen  Population  des  Tannenhahers 
{Nucifraga  caryocatactes)  bei  Praxmar,  Sellrain. 
Diplomarbeit,  Universitat  Innsbruck,  Austria. 

Kamil,  Alan  C;  Balda,  Russell  P.  1988.  Spatial  memory 
in  Clark's  nutcracker  {Nucifraga  columbiana).  Acta  XIX 
Congressus  Internat.  Ornithol.,  Ottawa  1986.  Vol.  II: 
2100-2111. 

Lanner,  Ronald  M.  1982.  Adaptations  of  whitebark  pine 
for  seed  dispersal  by  Clark's  nutcracker.  Canadian 
Journal  of  Forest  Research.  12(2):  391-402. 

Marzluff,  John  M.;  Balda,  Russ  P.  1992.  The  pinyon  jay: 
behavioral  ecology  of  a  colonial  and  cooperative  corvid. 
London:  Poyser.  317  p. 

Mattes,  Hermann  R.  1978.  Der  Tannenhaher  im  Engadin. 
Studien  zu  seiner  Okologie  und  Funktion  im  Arven- 
wald.  Miinstersche  Geographische  Arbeiten  2. 
Paderbom:  Schoning.  87  p. 

Mattes,  Hermann  R.;  Jenni,  Lukas.  1984.  Ortstreue  und 
Zugbewegungen  des  Tannenhahers  {Nucifraga  caryo- 
catactes) im  Alpenraima  und  am  Randecker  Maar/ 
Schwabische  Alb.  Ornithologischer  Beobachter.  81: 
303-315. 

Oberauer,  Eva.  1991.  Zur  Postembryonalentwicklimg  des 
Tannenhahers  {Nucifraga  caryocatactes  L.).  Beobach- 
tungen im  Alpenzoo  Innsbruck.  Diplomarbeit,  Uni- 
versitat Innsbruck,  Austria.  140  p. 

Reijmers,  N.  F.  1959.  Birds  of  the  cedar-pine  forests  of 
south-central  Siberia  and  their  role  in  the  life  of  the 
cedar-pine.  Trudy  biolog.  Inst,  sibirsk.  Otdel.  Akad. 
Nauk  SSSR.  5:  121-166. 

Rudat,  Volker;  Rudat,  Wolf.  1978.  Zum  Verhalten  von 
Tannenhaherfamilien  {Nucifraga  caryocatactes  L.)  in 
der  Zeit  vom  Ausfliegen  der  Jungen  bis  zu  deren 
Selbstandigwerden.  Zoologisches  Jahrbuch  fur 
Systematik.  105:  386-398. 

Saito,  Shin-ichiro.  1983.  On  the  relations  of  the  caching 
by  animals  on  the  seed  germination  of  Japanese  stone 
pine,  Pinus  pumila  Regel.  Bulletin  of  the  Shiretoko 
Museum.  5:  23-40. 

Swanberg,  Peer  O.  1951.  Food  storage,  territory  and  song 
in  the  thickbilled  nutcracker.  Acta  X  Congressus 
Internat.  Ornitholog.  Uppsala  1950:  545-554. 

Swanberg,  Peer  O.  1956.  Territory  in  the  thickbilled  nut- 
cracker {Nucifraga  caryocatactes).  Ibis.  98:  412-419. 


34 


Tomback,  Diana  F.  1977.  Foraging  strategies  of  Clark's 

nutcracker.  The  Living  Bird.  17:  123-161. 
Tomback,  Diana  F.  1982.  How  nutcrackers  find  their  seed 

stores.  The  Condor.  82:  10-19. 
Tomback,  Diana  F.;  Linhart,  Yan  B.  1988.  The  evolution 

of  bird-dispersed  pines.  Evolutionary  Ecology.  4: 

185-219. 

Vander  Wall,  Stephen  B.  1982.  An  experimental  analysis 
of  cache  recovery  in  Clark's  nutcracker.  Animal 
Behavior.  30:  84-94. 

Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1977.  Coadapta- 
tions  of  the  Clark's  nutcracker  and  the  pinyon  pine  for 


efficient  seed  har\'est  and  dispersal.  Ecological  Mono- 
graphs. 47(1):  27-37. 
Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1981.  Ecology 
and  evolution  of  food  storage  behavior  in  conifer- 
seed-caching  Corvids.  Zeitschrift  fiir  Tierpsychologie. 
56:  217-242. 

Woolfenden,  D.  E.;  Fitzpatrick,  W.-J.  1984,  The  Florida 
scrub  jay.  Princeton,  NJ:  Princeton  University  Press. 

Zink,  Gerhard.  1981.  Der  Zug  europaischer  Zugvogel.  3. 
Liefenmg.  Moggingen:  Vogelzug-Verlag. 


35 


ALLOZYME  POLYMORPHISM, 
HETEROZYGOSITY,  AND  MATING 
SYSTEM  OF  STONE  PINES 

Dmitri  V.  Politov 
Konstantin  V.  Krutovskii 


Abstract— Population  genetic  structure  and  the  mating  system  of 
stone  pines  (genus  Pinus,  subsection  Cembrae)  have  been  studied 
using  isozymes.  Observed  were  a  tendency  to  heterozygote  defi- 
ciency among  embryos  (obviously,  a  consequence  of  self-pollination 
with  the  estimated  fi-equency  of  0.11-0.30)  and  a  slight  heterozy- 
gote excess  in  adult  trees  (probably  caused  by  selection  against 
inbred  progeny  and  balanced  selection). 


Electrophoretic  analysis  of  allozjone  polymorphism  is 
one  of  the  most  useful  methods  for  study  of  genetic  and  evo- 
lutionary processes  in  natural  populations  and  taxonomic 
and  phylogenetic  relationships  of  forest  tree  species  (Adams 
and  others  1992;  Fineschi  and  others  1991;  MuUer-Starck 
and  Ziehe  1991).  Conifers  are  particularly  attractive,  from 
a  genetic  point  of  view,  because  of  their  haploid  megaga- 
metophyte  (endosperm).  Electrophoresis  of  both  haploid 
and  diploid  (embryo)  seed  tissues  permits  us  to  distinguish 
maternal  and  paternal  contributions  into  the  progeny  geno- 
type. This  makes  it  possible  to  estimate  allele  frequencies 
separately  in  both  male  and  female  gametes  pools  and  to 
study  segregating  ratios  directly  by  analyzing  endosperm 
haplotypes,  which  correspond  to  haplotypes  of  female  ga- 
metes. The  possibility  of  observing  meiotic  products  di- 
rectly through  study  of  haploid  endosperms  simplifies  analy- 
sis of  allozyme  inheritance,  genetic  Unkage  of  isozyme  loci, 
and  the  mating  system  of  conifers. 

In  contrast  to  many  widespread  European  and  North 
American  conifer  species,  stone  pines  have  been  studied 
using  isozymes  only  recently  (Groncharenko  and  others  1988; 
Krutovskii  and  Politov  1992;  Krutovskii  and  others  1987, 
1988, 1989, 1990;  Podogas  and  others  1991;  Politov  1989; 
PoUtov  and  Krutovskii  1990;  Politov  and  others  1989,  1992; 
Szmidt  1982).  Meanwhile,  the  need  for  such  data  is  appar- 
ent both  for  practical  forestry  applications  and  for  purposes 
of  genetic  diversity  conservation.  As  a  major  forest-forming 
tree  species,  stone  pines  determine,  to  a  large  extent,  the 
state  of  ecosystems  over  extensive  areas.  Their  genetic  vul- 
nerability thus  threatens  the  existence  of  a  unique  forest 
type. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Dmitri  V.  Politov  is  Research  Scientist  and  Konstantin  V.  Krutovskii 
is  Senior  Research  Scientist,  Laboratory  of  Population  Genetics,  Russian 
Academy  of  Sciences,  Gubkin  Str.  3, 117809  GSP-1  Moscow,  Russia. 


Four  of  the  five  stone  pine  species  traditionally  united 
in  subsection  Cembrae  of  genus  Pinus  (Critchfield  and 
Little  1966)  are  found  in  Russia  and  neighboring  countries: 
Siberian  stone  pine  (Pinus  sibirica  Du  Tour),  European  or 
Swiss  stone  pine  (P.  cembra  L.),  Korean  pine  {P.  koraiensis 
Siebold  et  Zucc),  and  mountain  or  Japanese  stone  pine 
(P.  pumila  [Pall.]  Kegel).  All  of  them  are  very  important 
economically. 

This  paper  reports  the  main  results  of  long-term  popula- 
tion genetic  studies  on  these  species  using  isozyme  analy- 
sis. Some  population  structure  parameters  also  have  been 
estimated  for  one  population  of  whitebark  pine  (P.  albicaulis 
Engelm.),  the  only  representative  of  stone  pines  in  the 
Western  Hemisphere. 

MATERIALS  AND  METHODS 

Sampling  Sites— Seeds  collected  in  natural  populations 
served  as  a  source  of  laboratory  material.  Pinus  sibirica 
seeds  were  collected  from  11  localities.  Samples  desig- 
nated as  Filin  Klyuch  (F),  Mutnaya  Rechka  (M),  Maljd 
Kebezh  (MK),  Sobach'ya  Rechka  (SR),  and  Listvyanka  (L) 
originate  from  the  Western  Sayan  (Ermakovskii  District, 
Krasnoyarsk  Territory);  Yayl'u  (YA)  from  the  Altfd  Moun- 
tains (Gomo-Altaiskaya  Autonomous  Region);  Smokotino 
(S),  Zorkal'tsevo-1  (Z-1),  and  Zorkal'tsevo-3  (Z-3)  from  Tom- 
skaya  Region;  Turukhan  (TU)  from  the  north  of  Krasnoy- 
arsk Territory  (Turukhanskii  District);  and  N'alino  (NYA) 
from  the  Khanty-Mansiiskii  Autonomous  District  (Tjoimen- 
skaya  Region).  Besides,  one  locality  of  P.  cembra,  Ust'- 
Chornaya  (UCH)  from  the  Eastern  Carpathians  (Ukraine); 
three  localities  of  P.  pumila  from  the  north  of  Kamchatka 
Peninsula  designated  as  Grustnyi  (GR),  Uttyveem  (UT), 
and  Kamenskoye  (K);  and  three  localities  of  P.  koraiensis 
from  Khabarovsk  Territory  designated  Malokhekhtsirskoye 
(MKH)  and  Khor  (KHO),  and  from  Primorskii  Territory 
(Sikhote-Alin'  (SA))  have  been  also  studied.  The  sample 
of  P.  albicaulis  originates  from  the  Livingstone  Falls  (LF) 
population  (southwestern  Alberta,  Canada). 

Electrophoresis  and  Data  Processing— No  less  than 
seven  endosperms  and  embryos  per  tree  have  been  studied 
using  starch  gel  electrophoresis.  Totally,  536  trees  of  five 
species  (from  11  to  75  trees  per  population)  and  more  than 
3,500  embryos  (7  to  20  per  tree),  as  well  as  two  bulked  sam- 
ples of  seeds  representing  a  large  number  of  trees  (TU  and 
NYA),  have  been  analyzed.  Specimen  preparation,  buffer 
systems  used,  genetic  interpretation  of  zymograms,  and  des- 
ignations of  £dlozymes,  alleles,  and  loci  have  been  described 
elsewhere  (Krutovskii  and  others  1987;  Politov  1989). 


36 


Genotypes  of  particular  trees  were  inferred  from  segre- 
gation of  alleles  among  their  endosperms.  Then,  allele  and 
genot5T)e  frequencies  in  samples  of  both  mature  trees  and 
embryos,  as  well  as  in  the  effective  pollen  pool  (paternal 
alleles  of  embryos),  were  estimated.  Embryo  tissue  elec- 
trophoresis does  not  give  high  resolution  for  all  zones  of 
activity  and,  therefore,  calculations  based  on  embryo  sam- 
ples have  included  a  smaller  nimiber  of  loci.  Besides,  some 
activity  zones  in  individual  species  could  not  be  reliably  in- 
terpreted or  were  invariant.  Because  of  this,  specific  sets 
of  loci  used  for  particular  purposes  in  these  species  may  dif- 
fer somewhat.  More  detailed  description  of  statistic  proce- 
dures can  be  fovmd  elsewhere  (Krutovskii  and  others  1989; 
Politov  1989;  PoHtov  and  Krutovskii  1990;  Politov  and  oth- 
ers 1989). 

GENETIC  CONTROL  OF  ISOZYMES 
AND  LINKAGE  OF  LOCI 

We  have  described  inheritance  of  11  isozyme  systems 
of  stone  pines  in  our  previous  publications  (Krutovskii  and 
others  1987;  Politov  1989)  and  foimd  that  they  are  controlled 
by  at  least  26  loci.  However,  only  21  most  reliable  loci  with 
unambiguous  genetic  control  have  been  included  in  this 
study  (table  1).  Segregation  ratios  of  allelic  variants  of  in- 
dividual allozjnne  loci  observed  among  endosperms  of  het- 
erozygous trees  correspond,  with  several  exceptions,  to  the 
expected  ones  (1:1). 

Analysis  of  joint  allelic  segregation  for  pairwise  loci  com- 
binations in  Pinus  sibirica  has  revealed  genetic  linkage 
of  loci  Adh-1  and  Lap-3  (with  the  mean  recombination  fi^e- 
quency  0.275),  Adh-1  and  Pgi-2  (0.301),  and  Fe-2  and  Lap-2 
(0.309).  These  loci  belong  to  two  linkage  groups:  Lap-3 — 
Adh-1— Pgi-2  and  Lap-2— Fe-2  (Politov  and  others  1989). 
Comparison  of  these  data  with  those  available  in  literature 
shows  that  stone  pines,  as  well  as  other  representatives  of 
the  family  Pi/iacme,  are  characterized  by  extreme  conser- 
vatism of  their  genome,  since  the  same  loci  combinations 
with  nearly  the  same  values  of  recombination  fi-equency 
have  been  described  not  only  in  the  closely  related  P.  albi- 
caulis  (Furnier  and  others  1986),  but  also  in  many  hard 
pines  (see,  for  example,  Szmidt  and  Muona  1989).  Data 
presented  here  are  apparently  the  first  attempt  of  genetic 
mapping  in  one  of  the  stone  pines.  However,  one  can  as- 
sume that  the  linkage  groups  revealed  in  P.  sibirica  will 
also  be  fotmd  in  other  related  species.  Information  of  this 
kind  could  be  useful  for  breeding  programs,  for  analysis 
of  association  of  genetic  markers  with  economically  impor- 
tant characters,  and  for  mating  system  parameters  estima- 
tion by  multilocus  methods. 

POPULATION  GENETIC  STRUCTURE 

Levels  of  Intrapopulation  Variability — ^According 
to  the  gene  diversity  parameters  estimated  in  stone  pine 
populations  using  isozjone  loci,  the  level  of  genetic  vari- 
ability appears  to  be  rather  high  (table  2).  There  are  al- 
most no  differences  between  observed  heterozygosities  and 
those  expected  fi'om  corresponding  Hardy- Weinberg  pro- 
portions. The  values  of  the  latter  for  individual  species 
are:  for  P.  sibirica— from  0.140  to  0.176,  P.  cembra— 0.109, 


Table  1 — Enzymes  and  loci  analyzed 

Total 
number 


Enzyme  (abbreviation) 

of  loci 

Scored  loci 

Alcohol  dehydrogenase  (ADH) 

2 

Adh-1 ,  Adh-2 

DIaphorase  (DIA) 

3 

Dia-1 ,  Dia-2 

Fluorescent  esterase  (FE) 

3 

Fe-2 

Glutamate  dehydrogenase  (GDH) 

1 

Gdh 

G 1  utam  atoxaloacetate 

3 

Got-1 ,  Got-2,  Got-3 

transaminase  (GOT) 

Isocitrate  dehydrogenase  (IDH) 

1 

Idh 

Leucine  aminopeptidase  (LAP) 

3 

LaD-2  LaD-3 

Malate  dehydrogenase  (MDH) 

4 

Mdh-1,  Mdh-2,  Mdh-3 

Mdh-4 

Phosphoglucose  isomerase  (PGI) 

2 

Pgi-2 

Phosphoglucomutase  (PGM) 

2 

Pgm-1,  Pgm-2 

Shikimate  dehydrogenase  (SKDH) 

2 

Skdh-1 ,  Skdh-2 

P.  pumila — from  0.239  to  0.258,  P.  koraiensis — from  0.113 
to  0.148,  P.  albicaulis— 0.204. 

Such  a  high  amoimt  of  intrapopulation  variability  is  char- 
acteristic of  conifers.  We  have  summarized  data  on  allozyme 
diversity  of  30  pine  species  using  available  references 
(Politov  and  others  1992).  The  mean  number  of  alleles  per 
locus  (A)  and  the  expected  heterozygosity  (Hg),  averaged  for 
all  pine  species,  equalled  2.08  and  0.169,  correspondingly. 
Comparable  values  have  been  found  for  spruces  (A  =  2.2, 
Hg  =  0.183),  larches  (A  =  1.6,  H^  =  0.091),  and  representa- 
tives of  other  less-studied  genera. 

Factors  supposedly  promoting  a  high  level  of  genetic  diver- 
sity in  stone  pine  populations  include  such  ones,  common 
for  forest  trees,  as  predominantly  outcrossing  mating  sys- 
tem, wide  distribution,  large  effective  size  of  population, 
and  high  longevity  (Hamrick  and  Godt  1989;  Ledig  1986). 
More  specific  evolutionary  factors,  such  as  drastic  reduc- 
tion of  population  size  ("bottleneck  effect"),  may  also  affect 
the  variability  level.  Although  the  ranges  of  stone  pines, 
both  in  Eurasia  and  in  North  America,  were  reduced  diu-- 
ing  the  Quaternary  glacials  (and  also  during  the  global 
warming),  preservation  of  extensive  refugiimis  permitted 
avoidance  of  substantial  genetic  depauperization.  This 
concerns  Pinus  cembra  to  a  much  less  degree  than  other 
stone  pines. 

According  to  some  evidence,  Swiss  stone  pine  imderwent 
a  sharp  reduction  in  its  range  and  population  size  during  the 
Holocene  caused  by  the  global  warming  (Bobrov  1978  and 
references  therein).  A  small  effective  number  is  typical  of 
most  populations  of  this  species  at  present  as  well.  Swiss 
stone  pine  exists  nowadays  in  the  Alps  and  the  Carpathians 
as  small  isolated  stands  in  which  processes  of  genetic  drift 
do  not  allow  aUozyine  variability  to  be  maintained  on  a  high 
level.  This  is  a  most  probable  cause  for  a  lower  level  of  het- 
erozygosity (0.109)  and  the  percentage  of  polymorphic  loci 
(26.3  percent)  of  P.  cembra,  as  compared  to  other  species. 
However,  it  should  be  noted,  that  the  only  sample  of  this  spe- 
cies studied  by  us  represents  a  small  marginal  stand  from 
the  eastern  Carpathians,  while  central  populations  may 
appear  to  be  more  variable.  The  limited  nimaber  of  loci 
studied  in  the  only  work  (Szmidt  1982)  dealing  with  allo- 
zyme polymorphism  in  Swiss  stone  pine  over  the  species 


37 


Table  2 — Levels  of  allozyme  variability  in  stone  pine  populations 


Species  and  Number  Number 


population  names 

of  loci 

of  trees 

P2 

Ho' 

He' 

Dim  1^  ^ihifS^o 
ninus  SiOiriCa 

Filin  Klyuch^  (F«) 

19 

53.0 

1.8 

47.4 

0.155 

0.152 

Mutnaya  Rechka  (M) 

19 

75.0 

1.9 

47.4 

.137 

.140 

Maiyi  Kebezh  (MK) 

19 

15.0 

1.8 

52.6 

.158 

.153 

Sobach'ya  Rechka  (SR) 

19 

12.0 

1.7 

42.1 

.145 

.173 

Listvyanka  (L) 

19 

41.0 

1.9 

42.1 

.163 

.151 

Yayl'u  (YA) 

19 

43.0 

1.9 

47.4 

.153 

.154 

Smokotino  (S) 

19 

43.0 

1.7 

42.1 

.168 

.161 

Zorkal'tsevo-I  (Z-1) 

19 

34.0 

1.7 

47.4 

.198 

.176 

Zorkartsevo-3  (Z-3) 

19 

34.0 

1.7 

47.4 

.184 

.165 

Mean 

19 

38.9 

1.8 

46.2 

.162 

.158 

UsfChornaya  (UCH) 

19 

15.0 

1.5 

26.3 

.128 

.109 

Pinus  pumila 

Grustnyi  (GR) 

20 

54.0 

2.3 

50.0 

.230 

.239 

Uttyveem  (UT) 

20 

55.0 

2.5 

60.0 

.243 

.251 

Kamenskoye  (K) 

20 

55.0 

2.3 

70.0 

.269 

.258 

Mean 

20 

54.7 

2.4 

60.0 

.247 

.249 

Pinus  koraiensis 

Malokhekhtsirskoye  (MKH) 

17 

30.0 

1.6 

47.1 

.147 

.132 

Khor  (KHO) 

17 

16.0 

1.6 

47.1 

.162 

.148 

OlKllOie-Mlin  ^OA; 

1  7 
1  / 

1  1  .u 

1  .O 

.Mo 

Mean 

17 

19.0 

1.6 

43.2 

.137 

.131 

Pinus  albicaulis 

Livingstone  Falls  (LF) 

16 

52.0 

1.9 

56.3 

.213 

.204 

'Mean  number  of  alleles  per  locus. 

^Mean  percent  of  loci  polymorphic  (locus  has  been  considered  polymorphic  if  the  frequency  of  the  most  com- 
mon allele  has  not  exceeded  95  percent). 
^Mean  heterozygosity  observed. 
"Mean  heterozygosity  expected. 
^See  text  for  origins  of  populations. 
^Abbreviated  denomination  of  population. 


range  does  not  allow  any  conclusions  to  be  made  about  the 
variability  level. 

Heterozygosity  of  Korean  pine  (mean  0.131)  is  slightly 
higher.  As  to  Pinus  pumila,  this  species  shows  the  highest 
values  of  H  among  stone  pines  (0.249),  close  to  the  maxi- 
mum ones  known  for  conifers  (Politov  and  others  1992). 
One  cannot  exclude  the  fact  that  such  a  high  variability 
is  very  important  for  adaptation  of  this  species  to  a  wide 
spectnmti  of  environmental  conditions,  including  those  of 
permafrost  regions,  and  drastic  seasonal  climatic  changes. 
Nearly  the  same  heterozygosity  level  (0.204)  was  found  in 
the  only  studied  population  of  whitebark  pine  from  the 
north  moimtain  area. 

Genotype  Distribution  in  Adult  Trees  and 
Embryos — Comparison  of  observed  genotype  distributions 
with  those  expected  from  the  corresponding  Hardy- Weinberg 
proportions  is  commonly  used  in  population  genetics  to  re- 
veal and  identify  various  factors  affecting  genetic  structure 
of  populations.  In  stone  pines  such  analysis  has  not  re- 
vealed pronounced  deviations  from  Hardy- Weinberg  equi- 
librium for  mature  trees  in  any  samples. 


Statistically  significant  excess  or  deficiency  of  heterozy- 
gotes  have  not  been  observed  for  allozyme  loci  studied  in 
stone  pine  populations.  Nevertheless,  the  sign  criterion 
showed  that  there  were  significantly  more  cases  of  a  slight 
heterozygote  excess  (Politov  and  others  1992),  which  may 
mean  existence  of  a  tendency  toward  systematic  excess  of 
heterozygotes  of  alloz3mie  loci  among  adult  trees.  Figure  1 
shows  mean  values  of  Wright's  fixation  index  (or  "inbreed- 
ing coefficient,"  F  =  l-H/Hg,  where  H^  and  Hg  are  mean 
observed  and  expected  heterozygosities  for  all  loci  studied 
in  the  species).  Predominance  of  slightly  negative  or  almost 
zero  F  values  indicates  absence  of  significant  heterozygote 
deficiency  in  the  reproductive  part  of  the  popvilations.  Spe- 
cies mean  values  were:  for  P.  sibirica,  -0.025;  for  P.  cembra, 
-0.174;  for  P.  koraiensis,  -0.046.  F  values  for  P.  pumila 
were  slightly  positive  in  two  samples  and  negative  in  the 
third  one  (mean  F  =  0.008). 

In  most  embryo  samples,  on  the  contrary,  a  slight  defi- 
ciency of  heterozygotes  was  regularly  observed  (Politov  and 
others  1992).  Figure  2  presents  mean  F  values  in  embryos 
and  adult  trees  calculated  for  the  same  loci.  In  embryos 


38 


0.1 
0.05 

F  0 

-0.05 
-0.1 
-0.15 


A. 
'1 

niri 

1 

ij 

1 

Populations:    F    M  MK  SR  L  YA  S  Z-1  Z-3  MEAN 


GR  UT  K      MEAN        MKH      SA  MEAN 


J  KHO 
I  


J 


P.  koraiensis 


P.  sibirica     P.  cembra   P.  pumila 

Figure  1 — Observed  fixation  index  ("inbreeding  co- 
efficient") values  (F)  in  samples  of  adult  trees.  All 
studied  loci  used.  (For  full  names  and  origins  of 
populations  see  text  and  table  2). 


the  values  are  positive  in  nearly  all  samples  (except  one), 
whereas  mature  trees  display  mainly  negative  F  values. 

Probable  Causes  of  Heterozygote  Deficiency — 

most  probable  cause  for  heterozygote  deficiency  at  early  de- 
velopmental stages  of  stone  pines  may  be  inbreeding.  The 
species  studied,  as  well  as  other  conifers,  have  a  mixed  mat- 
ing system  (Brown  and  others  1985;  Politov  and  Krutovskii 
1990  and  references  therein)  characterized  by  a  random 
outcrossing  with  a  certain  percentage  of  selfing.  Heterozy- 
gote deficiency  may  theoretically  be  caused  by  both  self- 
pollination  and  other  consanguineous  matings  (Ritland  and 
El-Kassaby  1985;  Shaw  and  others  1981).  One  cannot  also 
exclude  the  action  of  natural  selection  against  heterozy- 
gotes  at  a  certain  developmental  stage. 

Much  evidence  has  been  reported  demonstrating  hetero- 
zygote deficiency  among  embryos  in  populations  of  conifers 
(summarized  by  Bush  and  Smouse,  1992).  In  some  rare 
cases  deficiency  is  revealed  in  mature  trees.  The  authors 
explained  this  phenomenon  by  inbreeding  and,  particularly, 
by  selfing. 

Mating  System  Analysis — ^The  fact  that  selfing  is  most 
commonly  regarded  as  a  factor  causing  inbreeding  in  conifer 
species  has  good  reasons.  Coincidence,  in  general,  of  polli- 
nation with  pistil  receptivity  of  the  same  tree  and  absence  of 
effective  self-sterility  mechanisms  make  this  process  highly 
probable.  In  order  to  test  the  h3T)othesis  that  partial  selfing 
is  the  main  cause  of  heterozygote  deficiency  in  stone  pines, 
we  have  applied  methods  of  mating  system  parameters  es- 
timation based  on  allozyme  data  and  mixed  mating  system 
theory  (Brown  and  others  1985).  Use  of  isozymes  and  pow- 
erful mathematical  procedures  (Expectation-maximization 
method,  Newton-Raphson  method,  bootstrap,  etc.)  have 
made  it  possible  to  obtain  quantitative  estimates  of  selfing 
and  outcrossing  rates  in  natural  conifer  populations.  For 
different  species  percentage  of  selfing  varies  between  2  to  3 


and  20  to  30  percent  (see  Brown  and  others  1985;  Politov 
and  Krutovskii  1990,  for  review).  For  stone  pines  such 
estimates  have  not  been  made  until  recently.  Here  we 
present  data  obtained  on  three  stone  pine  species. 

The  results  of  mating  system  analysis  of  Siberian  stone 
pine  have  been  published  earlier  (Politov  and  Krutovskii 
1990).  If  we  consider  a  randomly  outcrossing  population 
at  eqmlibrium,  absence  of  pre-  and  postzygotical  selections, 
sexual  symmetry,  separated  generations,  etc.,  the  expected 
inbreeding  coefficient  Fg  can  be  derived  from  the  equation 
Fg  =  (l-^)/(l+^),  where  t  is  the  outcrossing  rate  estimated 
for  this  species.  In  our  study  the  expected  inbreeding  coef- 
ficient Fg  has  turned  out  to  be  very  close  to  the  actually 
observed  one  (F^)  [F„  =  iK~R^)fH^,  where  Hg  and  H^  are, 
correspondingly,  the  expected  from  Hardy- Weinberg  pro- 
portions and  the  directly  observed  heterozygosities]  in  em- 
bryo samples.  Thus,  selfing  (1-t)  whose  estimated  rate  is, 
on  the  average,  as  high  as  15  percent  (from  2  to  27  in  indi- 
vidual populations)  is  most  likely  to  be  one  of  the  main  fac- 
tors of  heterozygote  deficiency  occurrence  among  embryos 
of  this  species. 

Available  data  for  three  stone  pine  species  are  presented 
in  table  3.  Estimation  of  outcrossing  rates  was  made  based 
on  four  to  eight  polymorphic  loci  and  using  two  different 
computation  procedures — variants  of  both  single-  and  mtdti- 
locus  methods  (Ritland  and  El-Kassaby  1985;  Ritland  and 
Jain  1981).  Both  procedures  have  been  realized  by  Ritland 
(1990)  in  his  original  computer  program  MLT.  The  multi- 
locus  estimation  (which  has  more  detective  power  than  the 
single-locus  one)  gives  for  Pinus  sibirica  populations  values 
ranging  from  0.846  to  0.980  (mean  0.894),  and  for  P.  korai- 
ensis values  fi:'om  0.920  to  1.034  (mean  0.974).  The  out- 
crossing rate  foimd  in  P.  cembra  was  substantially  lower 


P.  sibirica 


P.  cembra  i_ 


Q  Adult  trees 
I  Embryos 


Figure  2 — Observed  fixation  index,  "inbreeding 
coefficient"  (F^),  values  in  samples  of  adult  trees 
(striped  bars)  and  embryos  (filled  bars).  Only 
eight  common  loci  used.  Fg  is  expected  inbreed- 
ing coefficient  [Fg  =  {^-tJ^{^+tJ].  S  is  selfing 
rate  (s  =  l-f^,  see  table  3).  (For  full  names  and 
origins  of  populations  see  text  and  table  2). 


P.  lioraiensis 


39 


Table  3 — Multilocus  (f  J  and  single-locus  {Q  estimates  of  outcrossing 
rate  in  stone  pine  populations 


Species  and         Number   Outcrossing  rate 

population  names      of  loci         t„  (s.e.)  (s.e.) 


Pinus  sibirica 

Filin  Klyuch-85^ 

8.0 

0.817 

(0.031) 

0.788 

(0.030) 

Filin  Klyuch-86 

8.0 

.892 

(.042) 

.884 

(.052) 

Mutnaya  Rechka-86 

8.0 

.980 

(.019) 

.959 

(.048) 

iviuiiidyct  nuoiirVd  0/ 

8.0 

.863 

\.\JOO) 

.0^0 

1  n'^fi^ 

Maiyi  Kebezh 

8.0 

.960 

(.072) 

.933 

(.060) 

Sobach'ya  Rechka 

8.0 

.846 

(.125) 

.834 

(.105) 

^lOlVyCII  IrVCl 

8.0 

.912 

(.031) 

fl71 

Yayl'u 

8.0 

.855 

(.045) 

.776 

(.043) 

Smokotino 

7.0 

.929 

(.033) 

.886 

(.037) 

Mean 

7.9 

.894 

(.057) 

.862 

(.054) 

Pinus  cembra 

Ust'-Chornaya 

4.0 

.686 

(.025) 

.707 

(.045) 

Pinus 

koraiensis 

Malokhekhtsirskoye 

6.0 

.967 

(.062) 

.914 

(.052) 

Khor 

7.0 

.920 

(.037) 

.929 

(.066) 

Sikhote-Alin' 

6.0 

1.034 

(.070) 

.964 

(.031) 

Mean 

6.3 

.974 

(.058) 

.936 

(.051) 

'Number  means  year  of  sampling  of  material;  see  text  for  origins  of 
populations. 


(0.686),  but  we  regard  this  value  only  as  a  preliminary  one 
because  of  the  limited  sample  size  and  number  of  loci  stud- 
ied. Nevertheless,  it  may  also  be  the  real  level  of  selfing, 
probably  due  to  the  limited  nimiber  of  trees  in  the  studied 
locality  and  a  low  stand  density. 

Shaw  and  others  (1981)  note  that  we  can  conclude  the  role 
of  inbreeding  components  other  than  selfing  per  se  in  con- 
sanguineous matings  (for  example,  mating  between  closely 
related  neighbor  trees  in  a  population  with  family  structure 
or  assortative/selective  pollination/fertilization  between 
genetically  similar  trees)  from  the  difference  between  the 
values  of  multi-  and  single-locus  estimates.  In  oiu*  study  of 
stone  pines  these  differences,  for  nearly  all  samples,  appesir 
to  be  negligible  (the  difference,  in  general,  does  not  exceed 
0.04).  This  implies  a  rather  weak  effect  of  mating  between 
relatives  or,  in  other  words,  absence  of  microsubdivision  of 
the  population. 

Fumier  and  others  (1987)  found  that  genetic  and  physical 
distances  between  clumps  of  Pinus  albicaulis  fail  to  corre- 
late. The  authors  related  this  fact  to  the  effective  "mixing" 
of  seeds  in  the  process  of  their  dispersal  by  the  Clark's  nut- 
cracker {Nucifraga  columbiana  Engelm).  We  can  expect  to 
observe  randomization  of  such  kind  (more  closely  spaced 
seedlings  or  clumps  are  not  more  related  ones)  in  any  other 
pine  species  depending  on  the  European  nutcracker  (AT.  cary- 
ocatactes  L.)  in  their  dispersal,  since  the  effectiveness  of  nut- 
crackers as  seed  dispersers  has  been  repeatedly  proved  by 
numerous  investigations  (Lanner  1980,  1990;  Tomback 
and  Linhart  1991). 

Selection  Against  Inbred  Progeny — What  may  cause 
"disappearance"  of  heterozygote  deficiency  observed  among 
embryos,  as  compared  to  adult  trees  of  the  population 


(figs.  1  and  2)?  This  can,  most  likely,  be  the  result  of  selec- 
tion against  inbred  progeny  formed  by  self-pollination,  as 
well  as  selection  in  favor  of  heterozygotes  (overdominance). 
Both  selections  may  increase  heterozygosity  up  to  the  equi- 
librium level,  but  only  as  a  result  of  negative  assortative 
fertilization  or  selection  in  favor  of  heterozygotes  (balancing 
selection)  can  heterozygosity  also  exceed  Hardy-Weinberg 
expectations.  This  process  can  be  detected  with  the  aid  of 
isozyme  gene  markers  if  they  are  either  involved  in  it  di- 
rectly or  £ire  in  gametic  linkage  disequilibrium  with  selec- 
tively significant  genes  (Crow  and  Kimura  1970). 

In  naturgJ  conifer  populations  heterozygosity  of  adult 
trees  is  often  higher  than  in  embryos  (see  Bush  and  Smouse 
1992;  Krutovskii  and  others  1988;  Politov  and  others  1992, 
for  review).  For  instance,  in  Pinus  radiata  (Plessas  and 
Strauss  1986)  heterozygosity  of  yoimg  trees  (17  to  20  years) 
exceeded  that  of  5-year  seedlings  and  in  the  latter  it  was 
higher  than  in  embryos.  Yazdani  and  others  (1985)  found 
higher  heterozygosity  of  young  £ind  adult  Scots  pine  trees 
as  compared  to  embryos;  differences  between  the  young  and 
the  adult  trees  were  slight  indicating  the  action  of  selec- 
tion exactly  at  early  stages.  For  stone  pines  we  can  suggest 
a  pEirticular  life  stage  of  natural  selection  acting  against 
homozygotes — ^when  groups  of  seedlings  begin  to  germinate 
out  of  nutcrackers'  caches  competing  for  resources.  It  is 
noteworthy  that  Eurasian  stone  pines  (as  compared  to  white- 
bark  pine)  display  multistem  forms  rarely,  which  may  be 
the  result  of  intensive  competition  between  seedlings  of  the 
same  cache. 

More  direct  evidence  was  obt£iined  in  experiments  carried 
out  by  Farris  and  Mitton  (1984),  who  studied  embryos  and 
seedlings  of  yellow  pine  {Pinus  ponderosa)  and  foimd  the 
increase  of  heterozygosity  with  age.  Muona  and  others 
(1987)  reveeded  heterozygosity  deficit  in  Scots  pine  embryos 
and  its  "disappearance"  in  1.5-year-old  seedlings. 

Inbred  progeny  elimination  at  early  developmental  stages 
in  conifers  is  also  supported  by  some  nonallozyme  data  (see 
for  example,  Koski  1973). 

CONCLUSIONS 

Study  of  alloz5rme  variability  in  stone  pines  has  shown 
the  genetic  diversity  tjrpical  to  widespread  and  abimdant 
conifer  species  with  continuous  ranges.  Differences  between 
stone  pines  can  be  explained  by  both  the  events  of  their 
recent  evolutionary  history  and  adaptive  significance  of 
the  allozyme  variation.  Heterozygote  deficiency  among  em- 
bryos and  slight  heterozygote  excess  in  mature  trees  are 
common  features  of  stone  pines,  as  well  as  other  conifers. 
A  substantial  part  of  self-pollinated  progeny  is  eUminated 
during  ontogenesis  by  natural  selection. 

ACKNOWLEDGMENTS 

We  are  greatly  thankful  to  G.  R.  Furnier,  V.  N.  Vorobjev, 
N.  A.  Vorobjeva,  L.  I.  Milyutin,  R.  T.  Gut,  and  P.  Raztmiov 
for  providing  seeds  and  aid  in  their  sampling,  to  Yu.  S. 
Belokon'  for  helpful  assistance  in  the  laboratory  analysis, 
to  Yu.  P.  Altukhov  for  constant  support  of  our  study  and 
very  useful  conmients  on  this  meinuscript,  and  to  K.  Ritland 
for  kindly  providing  us  with  the  computer  program  MLT. 


40 


We  also  appreciate  H.-R.  Gregorius's  invaluable  comments 
that  helped  to  substantially  improve  the  manuscript. 

REFERENCES 

Adams,  W.  T.;  Strauss,  S.  H.;  Copes,  D.  L.;  Griffin,  A.  R., 
eds.  1992.  Population  genetics  of  forest  trees:  Proceed- 
ings of  the  International  Symposiimi  on  Population  Ge- 
netics of  Forest  Trees;  1990  July  31-August  2;  Corvallis, 
OR,  U.S.A.  Kluwer  Academic  Publishers.  420  p.  (see  also 
New  Forests.  1992.  6.) 

Bobrov,  E.  G.  1978.  Forest-forming  conifers  of  the  USSR. 
Moscow:  Nauka.  190  p.  [In  Russian]. 

Brown,  A.  H.  D.;  Barrett,  S.  C.  H.;  Moran,  G.  F.  1985.  Mat- 
ing system  estimation  in  forest  trees:  Models,  methods 
and  meanings.  In:  Gregorius,  H.-R.,  ed.  Population  genet- 
ics in  forestry.  Lecture  Notes  in  Biomathematics.  Vol.  60. 
Berlin-Heidelberg-New  York-Tokyo:  Springer- Verlag: 
32-49. 

Bush,  R.  M.;  Smouse,  P.  E.  1992.  Evidence  for  the  adaptive 
significance  of  allozymes  in  forest  trees.  New  Forests.  6: 
179-196. 

Critchfield,  W.  B.;  Little,  E.  L.  1966.  Geographic  distribution 
of  the  pines  of  the  world.  Misc.  Publ.  991.  Washington, 
DC:  U.S.  Department  of  Agriculture.  99  p. 

Crow,  J.  F.;  Kimura,  M.  1970.  An  introduction  to  popula- 
tion genetics  theory.  New  York:  Harper  and  Row. 

Farris,  M.  A.;  Mitton,  J.  B.  1984.  Population  density,  out- 
crossing rate  and  heterozygote  superiority  in  ponderosa 
pine.  Evolution.  38(5):  1151-1154. 

Fineschi,  S.;  Cannata,  F.;  Malvolti,  M.  E.;  Hattemer,  H.  H., 
eds.  1991.  Biochemical  markers  in  the  population  genet- 
ics of  forest  trees:  Proceedings  of  the  Joint  Meeting  of  the 
Working  Parties  S2. 04-021  Population  Genetics  and  Eco- 
logical Genetics  and  S2. 04-05  Biochemical  Genetics  of 
the  International  Union  of  Forest  Research  Organizations 
(lUFRO);  October  1988;  Porano,  Italy.  SPB  The  Hague: 
Academic  PubHshing  bv.  251  p. 

Fumier,  G.  F.;  Knowles,  P.;  Aleksiuk,  M.  A.;  Dancik,  B.  P. 

1986.  Inheritance  and  linkage  of  allozymes  in  seed  tissue 
of  whitebark  pine.  Canadian  Journal  of  Genetics  and 
Cytology.  28(4):  601-604. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P. 

1987.  Effect  of  avian  seed  dispersal  on  the  genetic  struc- 
ture of  whitebark  pine  popidations.  Evolution.  41(3): 
607-612. 

Goncharenko,  G.  G.;  Padutov,  V.  E.;  Krutovskii,  K  V.; 
Podzharova,  Z.  S.;  Kirgizov,  N.  Ya.;  Politov,  D.  V.  1988. 
Level  of  genetic  variability  in  Pinus  sibirica  in  Altai. 
Doklady  Akademii  Nauk  SSSR.  299(1):  222-225.  [In 
Russian].  [English  translation  published  by  Pleniim 
PubHshing  Corp.  in  Doklady  Biological  Sciences;  1988 
September:  139-141]. 

Hamrick,  J.  L.;  Godt,  M.  J.  1989.  Allozyme  diversity  in 
plant  species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler, 
A.  L.;  Weir,  B.  S.,  eds.  Plant  population  genetics  breeding 
and  genetic  resources.  Sanderland,  MA:  Sinauer  Associ- 
ates: 43-63. 

Koski,  V.  1973.  On  self-fertilization,  genetic  load  and  sub- 
sequent inbreeding  in  some  conifers.  Comm.  Inst.  For, 
Fenn.  78(10):  1-40. 


Krutovskii,  K.  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1987. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  I.  (jenetic  control  of  isozyme  systems.  Genetika. 
23(12):  2216-2228.  [In  Russian].  [EngHsh  translation 
published  by  Plenimi  Publishing  Corp.  in  Soviet  Genet- 
ics; 1988  June:  1568-1576]. 

Krutovskii,  K  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1988. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  II.  Level  of  allozyme  variability  in  Western 
Sayan.  Genetika.  24(1):  118-124.  [In  Russian].  [EngHsh 
translation  published  by  Plenum  Publishing  Corp.  in 
Soviet  Genetics;  1988  July:  81-86]. 

Krutovskii,  K  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.;  Milyutin, 
L.  I.;  Kxiznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorob'ev,  V.  N.; 
Vorob'eva,  N.  A.  1989.  Genetic  variability  in  Siberian 
stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic  diversity 
and  differentiation  between  populations,  (jenetika.  25(11): 
2009-2032.  [In  Russian].  [EngHsh  translation  pubHshed 
by  Plenimi  Publishing  Corp.  in  Soviet  Cjenetics;  1990  May: 
1343-1362]. 

Krutovskn,  K.  V.;  Politov,  D.  V.;  Altukhov,  Yu.  P.  1990.  In- 
terspecific genetic  differentiation  of  Eurasian  stone  pines 
for  isoenzyme  loci.  Genetika.  26(4):  694-707.  [In  Russian]. 
[English  translation  published  by  Plenum  Publishing 
Corp.  in  Soviet  Genetics;  1990  October:  440-450]. 

Krutovskii,  K  V.;  Politov,  D.  V.  1992.  Study  of  intra-  and 
interspecific  genetic  differentiation  between  Eurasian 
cedar  pines  using  isozyme  loci  and  multidimensional 
analysis.  In:  Symposium  on  molecular  mechanisms  of 
genetic  processes:  proceedings.  Moscow:  Nauka  Publish- 
ers: 87-96.  [In  Russian]. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in  the 
ecology  and  evolution  of  limber  and  whitebark  pines.  In: 
6th  North  American  Forest  Biology  Workshop:  proceed- 
ings; 1980  August  11-13.  Edmonton,  AB:  University  of 
Alberta:  15-47. 

Lanner,  R.  M.  1990.  Biology,  taxonomy,  evolution,  and 
geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposimn  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermoimtain  Research  Station:  14-24. 

Ledig,  F.  T.  1986.  Heterozygosity,  heterosis,  and  fitness 
in  outbreeding  plants.  In:  Soule,  M.  E.,  ed.  Conservation 
biology:  the  science  of  scarcity  and  diversity:  77-104. 

Mvdler-Starck,  G.;  Ziehe,  M.,  eds.  1991.  Genetic  variation 
in  Eiiropean  populations  of  forest  trees:  proceedings  of 
the  workshop  on  genetic  variation  in  forest  tree  popula- 
tions in  Europe;  1990;  Gottingen,  Germany.  Frankfurt 
am  Main:  Sauerlander's  Verlag.  271  p. 

Muona,  O.;  Yazdani,  R.;  Rudin,  D.  1987.  Genetic  change 
between  life  stages  in  Pinus  syluestris:  allozjone  varia- 
tion in  seeds  and  planted  seedlings.  Silvae  Genetica. 
36(1):  39-42. 

Plessas,  M  E.;  Strauss,  S.  H.  1986.  Allozyme  differentiation 
among  populations,  stands,  and  cohorts  in  Monterey  Pine. 
Canadian  Journal  of  Forest  Research.  16(6):  1155-1164. 

Podogas,  A.  v.;  Shurkhal,  A.  V.;  Semerikov,  V.  L.;  Rakitskaya, 
T.  A.  1991.  Genetic  variabiHty  of  needle  tissue  enzymes 
of  Siberian  cedar  pine  (Pinus  sibirica  Du  Tour).  Genetika. 


41 


27(4):  695-703.  [In  Russian].  [English  translation  pub- 
lished by  Plenum  Publishing  Corp.  in  Soviet  Genetics; 
1991  October]. 
Politov,  D.  V.  1989.  Allozyme  polymorphism,  genetic  dif- 
ferentiation, and  mating  system  in  Siberian  stone  pine 
Pinus  sibirica  Du  Tour.  IVIoscow:  Institute  of  General 
Genetics,  Russian  Academy  of  Sciences.  190  p.  Thesis. 
[In  Russian]. 

Pohtov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1989. 
Genetic  variability  in  Siberian  stone  pine,  Pinus  sibirica 
Du  Tour.  III.  Linkage  relationships  among  isozyme  loci. 
Genetika.  25(9):  1606-1618.  [In  Russian].  [EngHsh  trans- 
lation published  by  Plenum  Publishing  Corp.  in  Soviet 
Genetics;  1990  IVIarch:  1053-1063]. 

PoHtov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variability 
in  Siberian  stone  pine,  Pinus  sibirica  Du  Toiu*.  V.  Analy- 
sis of  mating  system.  Genetika.  26(11):  2002-2011.  [In 
Russian].  [English  translation  published  by  Plenmn  Pub- 
hshing  Corp.  in  Soviet  Genetics;  1991  May:  1309-1316]. 

Pohtov,  D.  v.;  Krutovskii,  K.  V.;  Altukhov,  Yu.  P.  1992. 
Isozyme  loci  characteristics  of  gene  banks  of  populations 
of  cedar  pines.  Genetika.  28(1):  93-114.  [In  Russian]. 
[English  translation  published  by  Plenvun  Publishing 
Corp.  in  Soviet  Genetics;  1992  July:  76-95]. 

Ritland,  K.  1990.  A  series  of  FORTRAN  computer  programs 
for  estimating  plant  mating  system.  Journal  of  Heredity. 
81:  235-237. 


Ritland,  K.;  El-Kassaby,  Y.  A.  1985.  The  nature  of  inbreed- 
ing in  a  seed  orchard  of  Douglas  fir  as  shown  by  an  effi- 
cient multilocus  model.  Theoretical  and  Applied  Genetics. 
71:  375-384. 

Ritland,  K.;  Jain,  S.  1981.  A  model  for  the  estimation  of 
outcrossing  rate  and  gene  frequencies  using  n  indepen- 
dent loci.  Heredity.  47(1):  35-52. 

Shaw,  D.  v.;  Kahler,  A.  L.;  Allard,  R.  W.  1981.  A  multilocus 
estimator  of  mating  system  parameters  in  plant  popula- 
tions. Proceedings  of  the  National  Academy  of  Sciences 
(USA).  78:  1298-1302. 

Strauss,  S.  H.;  Bosquet,  J.;  Hipkins,  V.  D.;  Hong,  Y.-P.  1992. 
Biochemical  and  molecular  genetic  markers  in  biosys- 
tematic  studies  of  forest  trees.  New  Forests.  6:  125-158. 

Szmidt,  A.  E.  1982.  Genetic  variation  in  isolated  popula- 
tions of  stone  pine  {Pinus  cembra).  Silva  Fennica.  16(2): 
196-200. 

Szmidt,  A.  E.;  IMuona,  O.  1989.  Linkage  relationships  of 
allozyme  loci  in  Pinus  sylvestris.  Hereditas.  Ill:  91-97. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1991.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Biology.  4:  185-219. 

Yazdani,  R.;  Lindgren,  D.;  Rudin,  D.  1985.  Gene  dispersion 
and  selfing  frequency  in  a  seed-tree  stand  of  Pinus  sylves- 
tris (L.).  In:  Gregorius,  H.  R.,  ed.  Population  genetics  in 
forestry.  Lecture  Notes  in  Biomathematics.  Vol.  60.  Berlin- 
Heidelberg-New  York-Tokyo:  Springer- Verlag:  139-154. 


42 


GENETIC  POPULATION  STRUCTURE 
AND  GROWTH  FORM  DISTRIBUTION 
IN  BIRD-DISPERSED  PINES 

Diana  F.  Tomback 
WiUiam  S.  F.  Schuster 


Abstract — ^At  least  nine  species  of  the  20  wingless-seed  pines 
are  dependent  on  the  nutcrackers  {Nucifraga)  for  seed  dispersal. 
Seed  dispersal  by  nutcrackers  influences  several  aspects  of  popu- 
lation structure:  (1)  tree  dispersion  patterns  and  growth  form  dis- 
tribution, (2)  genetic  relationships  among  trees  within  popula- 
tions, and  (3)  genetic  diversity  among  populations.  Studies  to 
date  of  three  species,  whitebark  pine  {Pinus  albicaulis),  limber 
pine  {Pinus  flexilis),  and  Swiss  stone  pine  (Pinus  cembra),  pro- 
vide examples  of  intrapopulation  genetic  relationships  and  growth 
form  distribution;  other  recent  work  provides  some  information 
for  interpopulation  comparisons. 


of  parent  trees,  resulting  in  a  haphazard  distribution  of 
seeds  with  respect  to  seed  source  (Tomback  1978,  1982). 
Also,  nutcrackers  often  bury  more  than  one  seed  per  cache 
(mean  of  three  or  fotir  seeds  per  cache  for  Clark's  nut- 
cracker, Tomback  and  Linhart  1990),  and  the  seeds  may 
come  from  the  same  parent  tree  (Tomback  1988;  Tomback 
and  Knowles  1989). 

In  contrast,  most  wind-dispersed  seeds  fall  within  about 
120  m  of  parent  trees  and  are  dependent  on  prevailing 
winds  for  dispersal  (see,  for  example,  McCaughey  and  oth- 
ers 1986).  Updrafts  and  wind  patterns  associated  with 
storms  may  send  wind-dependent  seeds  over  longer  dis- 
tances and  in  other  directions,  and  seed-storing  rodents 
may  move  seeds  aroimd  locally  (Vander  Wall  1992).  How- 
ever, because  the  seeds  of  these  species  usually  move  rela- 
tively short  distances  and  establishment  is  often  in  small, 
disturbed  patches,  a  family  structure  (local  aggregations 
of  related  genotypes)  frequently  results  within  popula- 
tions (see,  for  example,  Knowles  1984;  Linhart  1989; 
Linhart  and  others  1981;  see  also  Fumier  and  others  1987; 
Tomback  and  Linhart  1990). 

These  different  seed-dispersal  modes,  bird  vs.  wind, 
may  lead  to  different  genetic  population  structures.  Ge- 
netic population  structure  resiilts  from  nonrandom  spatial 
distribution  of  genotypes  (Epperson  1990,  and  references 
therein).  Plant  populations  may  show  structure  on  differ- 
ent scales,  such  as  among  populations  and  within  popula- 
tions. The  common  methods  for  investigating  genetic 
population  structure  are  based  on  allozyme  analysis, 
comparing  populations  or  subpopxilations  by  means  of  ei- 
ther allele  frequencies,  Wright's      or  Nei's  Gg.^,  statistics, 
or  spatial  autocorrelation. 

Seed  dispersal  by  nutcrackers  influences  several  levels 
of  popiilation  structure  in  wingless-seed  pines.  Within 
populations,  it  affects  both  (1)  tree  dispersion  patterns 
(whether  individuals  are  clumped  vs.  spaced)  and,  conse- 
quently, growth  form  distribution  and  (2)  genetic  relation- 
ships among  trees.  Among  populations  it  affects  genetic 
diversity.  In  this  paper,  we  explore  the  specific  impacts 
of  nutcracker  seed-dispersal  behavior  on  each  of  these 
aspects  of  population  structure. 

GROWTH  FORM  DISTRffiUTION 


Of  the  20  species  of  pines  with  large,  wingless  seeds 
(collectively  considered  "stone  pines"),  nine  are  known 
dependents  on  the  nutcrackers,  Nucifraga  (Corvidae),  for 
seed  dispersal  (Tomback  and  Linhart  1990  and  references 
therein).  Five  of  these  bird-pines  are  in  the  Subsection 
Cembrae,  two  in  Strobi,  and  two  in  Parry  a;  for  classifica- 
tion see  Critchfield  and  Little  (1966).  For  several  species 
of  these  bird-dependent  pines,  other  corvids,  squirrels,  or 
other  rodents  may  also  make  varying  contributions  to  tree 
estabHshment  (see,  for  example,  Hayashida  1989;  Ligon 
1978;  Vander  Wall  1992),  but  the  factor  in  common  to  all 
dispersers  is  the  tendency  to  remove  seeds  from  the  vicin- 
ity of  parent  trees  and  "scatterhoard"  them  (biuy  small 
clusters  of  seeds  in  subterranean  caches  for  future  use). 
Clark's  nutcrackers  (A^.  columbiana)  bury  one  to  15  or 
more  seeds  per  cache  at  distances  of  several  meters  to 
22  km  from  parent  trees  (Hutchins  and  Lanner  1982; 
Tomback  1978,  1982;  Vander  Wall  and  Balda  1977).  Eur- 
asian nutcrackers  (A^.  caryocatactes)  bury  up  to  24  seeds 
per  cache  and  travel  similar  distances  (see,  for  example, 
Bibikov  1948;  Mattes  1982;  Mezhenny  1961;  Turcek 
1966). 

Thus,  nutcrackers  create  a  highly  impredictable  seed 
shadow,  with  seedhng  establishment  oft^n  occurring  sev- 
eral kilometers  from  parent  trees  (Tomback  and  others 
1990).  Nutcracker-dispersed  seeds  may  end  up  in  small, 
isolated  areas,  particularly  following  disturbances  such 
as  fires.  In  large,  open  or  disturbed  areas,  different  nut- 
crackers may  cache  seeds  from  several  difi"erent  stands 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Diana  F.  Tomback  is  Associate  Professor  of  Ecology,  Department  of 
Biology  and  Center  for  Environmental  Sciences,  University  of  Colorado 
at  Denver,  Denver,  CO,  80217-3364.  William  S.  F.  Schuster  is  Forest  Di- 
rector, Black  Rock  Forest  Consortium,  Box  483,  Cornwall,  NY  12518. 


Research  to  date  on  three  species  of  bird-dispersed 
pines — whitebark  pine  (Pinus  albicaulis)  (Fumier  and 
others  1987;  Linhart  and  Tomback  1985),  limber  pine 
(P.  flexilis)  (Carsey  and  Tomback  1992;  Linhart  and 
Tomback  1985;  Schuster  and  Mitton  1991),  and  Swiss 
stone  pine  (P.  cembra)  (Tomback  and  others  1993) — 


43 


indicates  that  trees  within  populations  occur  in  three 
different  growth  forms.  The  single-trunk  growth  form 
is  typical  of  conifers  and  consists  of  a  single  genet  (indi- 
vidual). Also  composed  of  a  single  genet,  the  multi-trunk 
growth  form  has  two  or  more  trunks  either  contiguous 
at  the  base  or  fused  at  the  base  or  part  way  up  the  bole. 
Multi-trunk  trees  may  be  the  consequence  of  a  tendency 
toward  side-branching  and  lack  of  apical  dominance,  par- 
ticularly under  stressful  environmental  conditions,  such 
as  unstable  soils,  aridity,  or  mechanical  damage  to  the 
leader  (Schuster  and  Mitton  1991;  Weaver  and  Jacobs 
1990).  Although  the  tree  cluster  growth  form  is  also 
composed  of  two  or  more  contiguous  or  fused  trunks,  the 
trunks  consist  of  different  genets  (individuals).  This 
growth  form  results  from  the  germination  and  survival 
of  more  than  one  seedling  from  a  seed  cache  (growth  form 
terminology  from  Tomback  and  others  1990). 

The  latter  two  growth  forms  have  similar  morphologies 
and  thus  cannot  be  distinguished  with  certainty  without 
genetic  analysis  (Schuster  and  Mitton  1991;  Tomback  and 
others  1993)  (figs.  1,  2).  Consequently,  in  field  surveys  of 
frequency  distribution  of  different  growth  forms,  we  com- 
bined the  multi-trunk  and  tree  cluster  forms  together  in 


Figure  1 — Swiss  stone  pine  {Pinus  cembra)  tree 
cluster.  Protein  electrophoresis  indicated  that  each 
trunk  is  a  different  genet. 


Figure  2 — Swiss  stone  pine  {Pinus  cembra) 
multi-trunk  tree.  Protein  electrophoresis  indi- 
cated that  both  trunks  had  the  same  genotype. 


the  category  "tree  clumps"  (term  from  Lanner  1980;  the 
same  as  "multi-trunk  trees  of  unknown  origin,"  Tomback 
and  others  1990).  For  reporting  the  results  of  field  sur- 
veys of  tree  clump  frequency  in  populations,  we  use  the 
term  "tree  site"  to  refer  to  the  small  area  of  ground  sup- 
porting either  a  single-trunk  tree,  a  multi-trimk  tree,  or 
a  tree  cluster. 

Overall,  tree  clumps  accounted  for  19  to  90  percent  of 
the  tree  sites  surveyed  in  three  species  of  bird-dispersed 
pines  (whitebark,  limber,  and  Swiss  stone)  (table  1). 
Whitebark  pine  in  the  two  stands  surveyed  had  the  high- 
est percentage  of  tree  sites  occupied  by  clumps.  In  the 
two  studies  of  limber  pine  combined,  the  percentage  of 
tree  sites  with  clvmips  ranged  from  24  to  48;  the  latter 
population  occurred  on  a  steep,  southeast-facing  slope 
with  gravelly  soil  at  3,310  m  elevation.  Carsey  and 
Tomback  (1992)  also  examined  17  tree  sites  in  a  stand 
of  krummholz  limber  pine  above  timberline  (3,460  m  eleva- 
tion) and  found  clumps  at  all  tree  sites.  For  Swiss  stone 
pine,  clumps  were  found  at  19  to  30  percent  of  the  tree 
sites  and  up  to  45  percent  of  seedling  sites.  In  the  Piz  da 
Staz  study  area,  the  highest  percentage  of  Swiss  stone 
pine  clumps  occurred  at  higher  elevations,  but  this  did 
not  hold  for  the  Muottas  Muragl  study  area. 


44 


Table  1 — Percentage  of  tree  clumps  (combined  multi-trunk  trees  and  tree  clusters)  in  surveyed  populations  of  three  species 
of  bird-dispersed  pines 


Species 

Maximum 

Percent 

and 

Tree 

trunks  per 

sites 

location 

Elevation 

sites 

clump 

with  clumps 

References 

Meters 

Swiss  stone  pine 

Switzerland 

1,835 

406 

6 

21 

Tomback  and 

2,070 

351 

6 

30 

others  1 993 

1 ,835 

96 

4 

29 

2,125 

43 

5 

19 

Whitebark  pine 

California 

2,850 

50 

5+ 

90 

Tomback  and 

Montana 

2,730 

50 

5+ 

58 

Linhart  1990 

Limber  pine 

Colorado 

1,650 

361 

7 

30 

Schuster  and 

Mitton  1991 

Colorado 

2,585 

105 

3 

35 

Carsey  and 

2,810 

135 

4 

24 

Tomback  1992 

3,310 

140 

4 

48 

The  preferred  laboratory  method  for  distinguishing  tree  stone  pine  clumps  sampled  in  one  stand,  70  percent  were 
clusters  from  multi-tnmk  trees  has  been  starch  gel  pro-  tree  clusters  (Tomback  and  others  1993).  Remarkably, 

tein  electrophoresis  to  assay  the  genotype  of  seed  (Fiimier  an  examination  of  the  number  of  gene  loci  used  for  each 
and  others  1987)  or  foliage  tissue  obtained  from  each  study  in  table  2  shows  no  relationship  with  the  percent- 

tnmk  of  a  tree  clump  (Carsey  and  Tomback  1992;  Linhart         age  of  clusters  determined  for  each  sample  (Spearman 
and  Tomback  1985;  Schuster  and  Mitton  1991;  Tomback  rank  correlation,    =  -0.406,  n  =  12,  NS).  With  the  excep- 

and  others  1993;  for  review  of  technique,  see  Kephart  tion  of  the  Umber  pine  clumps  analyzed  by  Schuster  and 

1990).  Those  clvunps  that  have  two  or  more  trimks  with  Mitton  (1991),  which  represented  all  the  climips  in  the 

different  genotypes  are  classified  as  tree  clusters.  Unfor-  popialation,  the  clvmips  examined  in  the  studies  in  table  2 

timately,  this  technique  may  underestimate  slightly  the 
nimibers  of  tree  clusters,  because  seeds  within  nutcracker 
caches  tend  to  be  genetic  relatives  and  thus  may  inherit 
similar  genotypes  (Carsey  and  Tomback  1992;  Schuster 
and  Mitton  1991).  In  a  few  cases  tree  cluster  members 
may  possess  identical  genotypes  at  the  gene  loci  exam- 
ined, so  the  clusters  are  misclassified  as  multi-trunk 
(single-genotype)  trees.  The  proced\ires  for  calculating 
the  probability  of  such  errors  are  presented  in  Schuster 
and  Mitton  (1991). 

For  studies  of  whitebark  pine,  hmber  pine,  and  Swiss 
stone  pine  combined,  genetic  analysis  indicated  that  18 
to  100  percent  of  the  tree  clumps  sampled  were  tree  clus- 
ters (table  2).  The  low  value,  18  percent,  was  obtained  by 
Schuster  and  Mitton  (1991)  for  a  limber  pine  population 
in  the  Pawnee  National  Grasslands,  at  the  periphery  of 
the  range  of  both  the  pine  and  Clark's  nutcracker.  The 
high  values  of  100  percent  for  limber  pine  and  83  percent 
for  whitebark  pine  were  based  on  small  sample  sizes 
in  =  six  clumps  in  both  cases,  Linhart  and  Tomback  1985) 
Filmier  and  others  (1987)  found  that  58  and  70  percent, 
respectively,  of  the  whitebark  pine  clumps  they  sampled 
in  two  stands  contained  more  than  one  genotype;  they  did 
not  sample  all  trunks  of  some  clumps,  so  these  percent- 
ages are  conservative.  Clumps  in  four  limber  pine  stands 
were  studied  by  Carsey  and  Tomback  (1992),  with  per- 
centages of  clusters  ranging  from  24  percent  above  tim- 
berhne  to  81  percent  at  2,810  m  elevation.  Of  the  Swiss 


Table  2 — Percentage  of  tree  clumps  (combined  multi-trunk  trees 
and  tree  clusters)  found  to  be  tree  clusters  (growth  form 
with  more  than  one  genet)  in  three  species  of  bird- 
dispersed  pines 


Species 

Clumps 
examined 

Gene 
loci 
used 

Percent 
clusters 

References 

Swiss  stone 

23 

4 

70 

Tomback  and  others 

pine 

1993 

Whitebark  pine 

6 

4 

83 

Linhart  and  Tomback 

1985 

12 

11 

58 

Furnier  and  others 

23 

11 

70 

1987 

Limber  pine 

7 

2 

57 

Linhart  and  Tomback 

6 

4 

83 

1985 

6 

4 

100 

108 

10 

18 

Schuster  and  Mitton 

1991 

18 

7-9 

56 

Carsey  and  Tomback 

21 

7-9 

81 

1992 

18 

7-9 

44 

17 

7-9 

24 

45 


represented  a  sample  of  the  total  number  of  climaps  present 
and  were  sometimes  not  selected  entirely  at  random.  For 
the  studies  by  Furnier  and  others  (1987)  and  Carsey  and 
Tomback  (1992),  only  reproductively  mature  tree  clumps 
were  sampled;  for  the  studies  by  Linhart  and  Tomback 
(1985)  and  Tomback  and  others  (1993),  clumps  were  usu- 
ally sampled  as  encountered,  but  an  effort  was  made  to 
include  some  seedlings  and  young  trees. 

Only  three  studies  permit  an  estimate  of  the  overall 
percentage  occurrence  of  tree  clusters  among  tree  sites 
in  populations  (table  3).  Limber  pine  tree  clusters  in  the 
Pawnee  National  Grasslands  population  occurred  at  only 
5  percent  of  the  total  tree  sites.  For  limber  pine  in  the 
Front  Range,  clusters  occurred  at  similar  percentages  of 
tree  sites,  19  to  21,  despite  differences  in  elevation.  Clus- 
ters of  Swiss  stone  pine  occupied  21  percent  of  the  tree 
sites  in  the  upper  subalpine  population. 

Apparently,  tree  clusters  typically  occupy  a  substantial 
proportion  of  the  tree  sites  in  each  population  of  whitebark, 
limber,  and  Swiss  stone  pine.  Moreover,  the  percentage 
of  individual  genets  in  a  population  that  are  found  in  clus- 
ters is  much  higher  than  the  proportion  of  tree  sites  sup- 
porting clusters.  Based  on  the  previous  studies,  12  to  40 
percent  of  the  individuals  in  populations  of  bird-dispersed 
pines  occur  in  a  highly  clumped  spatial  pattern,  primarily 
as  a  consequence  of  seed  dispersal  by  nutcrackers. 

If  stress  plays  a  role  in  the  growth  of  the  multi-trunk 
form  from  a  single  genet,  as  proposed  by  Schuster  and 
Mitton  (1991)  and  Weaver  and  Jacobs  (1990),  then  we 
might  predict  a  relatively  higher  percentage  of  multi- 
trunk  trees  in  more  stressful  environments.  In  fact,  about 
25  percent  of  the  limber  pine  tree  sites  in  the  Pawnee 
National  Grasslands  population  support  multi-trunk  trees 
(from  tables  1  and  2).  This  area  is  at  lower  elevations 
(1,650  m)  on  the  plains  east  of  the  Rocky  Mountains,  at 
the  eastern  boundary  of  the  limber  pine  range,  and  trees 
may  be  subjected  to  stress  factors  such  as  grazing,  fire, 
unstable  slopes,  heat,  and  extreme  aridity  (Lepper  1974; 
McNaughton  1984;  Schuster  and  Mitton  1991).  The  high- 
est estimated  percentages  of  multi-trunk  limber  pine  trees 
in  the  Front  Range  populations  studied  by  Carsey  and 
Tomback  (1992)  occurred  in  the  upper  subalpine  popula- 
tion (27  percent,  elevation  3,310  m)  and  in  the  population 
above  timberline  (76  percent,  elevation  3,460  m),  where 
winds  and  accompanying  desiccation  are  extreme  and  the 


Table  3 — Estimated  percentage  of  tree  sites  tliat  support  clusters  of 
individuals  in  populations  of  two  species  of  bird-dispersed 
pines 


Table  4 — Estimated  mean  and  standard  error  of  relatedness  of 
individuals  within  and  between  limber  pine  tree  clusters 


Percent 

Species       clusters  Elevation 


References 


Meters 

Swiss  stone  pine    21    2,050  to  2,240    Tomback  and  others  1 993 


Limber  pine 


5 
19 
19 
21 


1,650 
2,585 
2,810 
3,310 


Schuster  and  Mitton  1 991 
Carsey  and  Tomback  1 992 


Mean    Standard  error 


References 


Within  tree  clusters 

0.19  0.10  Schuster  and  Mitton  1 991 

.43  .13  Carsey  and  Tomback  1 992 

Between  tree  clusters 

.01  .04  Carsey  and  Tomback  1 992 


soil  is  gravelly  and  unstable.  We  also  calculated  that  only 
9  percent  of  the  tree  sites  of  the  upper  subalpine  popula- 
tion of  Swiss  stone  pine  supported  multi-trunk  trees.  In 
contrast  to  the  former  study  areas,  the  Swiss  stone  pine 
tree  sites  are  not  typically  characterized  by  aridity  and 
poor  soils,  which  may  explain  the  relatively  low  frequency 
of  multi-trunk  trees  (Tomback  1988). 

We  should  note  here  that  tree  clusters  are  not  known 
to  occur  in  Colorado  pinyon  pine  (Pinus  edulis),  which  is 
dispersed  by  both  Clark's  nutcrackers  and  pinyon  jays 
(Gymnorhinus  cyanocephalus).  Vander  Wall  and  Balda 
(1977)  report  that  only  one  seed  from  each  cache  grows 
to  maturity. 

GENETIC  RELATIONSHIPS 

Relationships  Within  and  Between  Tree 
Clusters — Because  nutcrackers  tend  to  harvest  and 
pouch  several  seeds  from  the  same  pine  cone  and  visit 
more  than  one  cone  from  the  same  tree  or  genet  (if  a  clus- 
ter) (Tomback  1988;  Tomback  and  Knowles  1989),  they 
tend  to  place  sibling  seeds  in  the  same  seed  cache.  Dur- 
ing a  bumper  whitebark  pine  cone  crop  (1989)  in  Yellow- 
stone National  Park,  Tomback  and  Knowles  (1989)  ob- 
served nutcrackers  taking  entire  pouchloads  of  28  to  97 
seeds  {x  =  53  seeds,  n  =  8)  from  a  single  cone  or  individual 
tree  (genet)  or  trimk. 

One  consequence  of  this  behavior  is  the  tendency  for 
individuals  in  tree  clusters  to  be  genetic  relatives — either 
half-siblings,  siblings,  or  the  products  of  selflng.  Furnier 
and  others  (1987)  found  that  individuals  within  tree  clus- 
ters of  two  populations  were  genetically  more  similar 
(smaller  genetic  distance)  than  individuals  in  different 
tree  clusters.  Individuals  within  clusters  differed  by  a 
mean  of  1.66  alleles;  individuals  of  different  clusters  dif- 
fered by  a  mean  of  5.82  alleles  or  greater  (of  11  gene  loci). 
Two  studies  examined  genetic  relatedness  within  limber 
pine  clusters  using  the  Quellar  and  Goodnight  (1989)  esti- 
mator (table  4):  For  tree  clusters  in  the  Pawnee  National 
Grasslands  population,  Schuster  and  Mitton  (1991)  deter- 
mined that  individuals  were  related  on  average  as  half- 
siblings.  In  addition,  they  determined  that  cluster  mem- 
bers that  had  fused  together  at  the  base  were  more  closely 
related  on  average  (r  =  0.35)  than  cluster  members  that 
had  not  fused  (r  =  0.05).  Carsey  and  Tomback  (1992)  cal- 
culated that  cluster  members  in  their  upper  subalpine 
population  (3,310  m  elevation)  were  related  on  average 
as  between  half  and  full  siblings;  members  of  different 
clusters  were  unrelated  on  average  (table  4). 


46 


Possible  Biological  Implications  of  Tree 
Clustering — Despite  the  fact  that  tree  clusters  occiir  com- 
monly in  populations  of  bird-dispersed  pines  and  include 
a  large  proportion  of  the  genets  in  the  population,  little  is 
known  of  the  biology  of  this  growth  form  (Tomback  and 
others  1993).  Several  papers  have  speculated  about  the 
potential  advantages  and  disadvantages  to  individuals 
growing  in  clusters  (Schuster  and  Mitton  1991;  Tomback 
and  Linhart  1990;  Tomback  and  others  1993). 

Individuals  in  clusters  may  have  higher  survivorship 
than  single  genets  vmder  certain  environmental  condi- 
tions because  of  morphological  or  physiological  benefits 
(Bullock  1981;  Keeley  1988;  Mattes  1982).  A  sturdy,  well- 
anchored  structure  might  better  resist  the  strong  winds 
and  mechanical  damage  of  higher  elevation  environments. 
Root  grafting  may  also  occur  among  cluster  members,  par- 
ticularly relatives  (Schuster  and  Mitton  1991),  resulting 
in  efficient  water  and  nutrient  acquisition  (Holtmeier 
1986).  In  addition,  since  nutcrackers  may  establish  pines 
at  some  distance  from  conspecifics,  for  example,  in  burned 
areas  or  disturbed  patches,  individuals  in  tree  clusters 
may  cross-pollinate  and  thus  avoid  selfing  (Tomback  and 
Linhart  1990).  This  might  produce  a  high  proportion  of 
sound  seeds,  particularly  if  clusters  consisted  of  both  rela- 
tives and  nonrelatives.  Finally,  both  Schuster  and  Mitton 
(1991)  and  Tomback  and  Linhart  (1990)  point  out  that  the 
occurrence  of  kin  groups  of  bird-dispersed  pines  may  re- 
sult in  kin  selection  as  an  evolutionary  force.  Possible 
consequences  for  cluster  members  may  include  some  de- 
crease in  competition  or  facilitation  of  rootgrafting. 

With  respect  to  the  potential  disadvantages  to  indivi- 
duals growing  in  clusters,  one  area  of  possible  fitness  loss 
is  reproductive  output.  For  limber  pine  trees  in  three 
stands  in  the  Colorado  Front  Range  (same  populations 
studied  by  Carsey  and  Tomback  1992),  cluster  members 
produced  fewer  male  and  female  cones  than  did  single- 
tnmk  trees  during  each  of  3  years  (Feldman  1991; 
Feldman  and  Tomback  1991).  Reduction  of  reproductive 
output  may  result  from  competition  among  cluster  mem- 
bers for  moisture,  nutrients,  and  space.  These  effects  may 
be  exacerbated  for  limber  pine,  which  typically  grows  on 
windy,  xeric  sites  with  poor  soils  in  the  Colorado  Front 
Range  (Feldman  1991;  Feldman  and  Tomback  1991). 
Feldman  (1991)  and  Feldman  and  Tomback  (1991)  noted 
that  trees  in  clusters  had  smaller  diameters  than  did 
similarly  aged  single-trunk  trees,  supporting  the  possibil- 
ity of  competition. 

Another  possible  disadvantage  to  cluster  members  is 
that  their  seeds  may  be  more  inbred  than  seeds  produced 
by  single-genet  trees,  because  of  pollen  exchange  within 
clusters  (Furnier  and  others  1987;  Tomback  and  Linhart 
1990).  PoUtov  and  Krutovskii  (1990)  provide  some  evi- 
dence for  this  in  Siberian  stone  pine  (Pinus  sibirica), 
which  is  dispersed  by  the  Eurasian  nutcracker.  Examin- 
ing the  mating  system  of  Siberian  stone  pine  in  seven 
populations,  they  noted  an  excess  of  homozygotes  among 
seed  embryos  in  the  majority  of  their  samples.  They  at- 
tributed this  to  a  relatively  high  rate  of  self-pollination, 
but  pollination  among  related  cluster  members  would 
have  a  similar  effect. 


Genetic  Spatial  Structure  Within  Popvilations — 

Several  studies  indicate  that  in  many  wind-dispersed 
conifers  there  is  family  structure  within  popiilations 
(local  aggregations  of  genetic  relatives)  (see,  for  example, 
Knowles  1984;  Linhart  1989;  Linhart  and  others  1981). 
This  is  the  consequence  of  seed  dispersal  typically  re- 
stricted to  short  distances  (Ftunier  and  others  1987; 
Tomback  and  Linhart  1990),  and  the  availability  of  local, 
disturbed  patches  for  colonization  (see,  for  example, 
Linhart  and  others  1981).  In  contrast,  Furnier  and  others 
(1987)  did  not  find  any  evidence  of  family  structure 
among  the  neighboring  tree  clusters  in  two  whitebark 
pine  populations  (although  cluster  members  tended  to 
be  related).  They  attributed  this  lack  of  family  structiire 
with  respect  to  geographic  distance  to  the  seed-caching 
behavior  of  Clark's  nutcrackers.  Since  Clark's  nutcrack- 
ers do  not  defend  caching  territories,  several  different 
nutcrackers  may  cache  in  one  area,  often  bringing  seeds 
in  from  different  stands  of  trees  (Tomback  1978,  1982). 
This  probably  results  in  a  haphazard  distribution  of  re- 
lated seed  caches  (Furnier  and  others  1987;  Tomback 
and  Linhart  1990).  The  Eurasian  nutcracker  in  Sweden 
(Swanberg  1956)  and  in  the  Alps  (Mattes  1982)  stores 
many  of  its  caches  within  territories  defended  by  pairs. 
However,  the  seeds  cached  in  these  territories  are  both 
from  local  and  more  distant  parent  trees  (up  to  15  km. 
Mattes  1982),  so  there  is  also  considerable  mixing  in  any 
area.  PoHtov  and  Krutovskii  (1990)  examined  seven  popu- 
lations of  Siberian  stone  pine  and  also  foimd  no  evidence 
for  intrapopulation  substructure. 

GENETIC  DIVERSITY 

Generally  speaking,  conifers  show  less  genetic  diversity 
among  popiilations  (mean      =  0.068)  than  do  other  kinds 
of  plants  (mean  Gg^  =  0.250)  (Hamrick  and  Godt  1990). 
After  an  extensive  review  of  the  plant  literature,  Hamrick 
and  Godt  (1990)  concluded  that  animal-dispersed  species 
have  more  differentiated  populations  (higher  Gg^  values) 
than  do  wind-dispersed  species.  But,  seed  dispersal  by 
nutcrackers,  which  occurs  routinely  over  large  distances, 
might  actually  result  in  higher  levels  of  gene  flow  than 
from  seed  dispersal  by  wind  and  thus  lower  levels  of  popu- 
lation differentiation. 

Few  studies  have  examined  interpopulation  variation  in 
bird-dispersed  pines.  Schuster  and  others  (1989)  geneti- 
cally analyzed  two  limber  pine  populations  at  extreme 
ends  of  a  1,700-m  elevational  transect  in  the  Colorado 
Front  Range.  They  determined  that  pollen  flow  along  this 
gradient  was  impeded  by  pollination  phenology;  popula- 
tions separated  by  more  than  400  m  elevation  did  not 
have  overlapping  pollination  periods.  For  the  populations 
at  each  end  of  the  transect,  allele  frequencies  differed  sig- 
nificantly at  eight  of  the  10  loci  studied.  However,  the 
calculated  nimiber  of  migrants  exchanged  between  the 
populations  per  generation  (A^^  =  11.1)  was  well  above  the 
level  required  to  overcome  genetic  drift.  Given  the  very 
different  environmental  conditions  of  the  two  populations, 
Schuster  and  others  (1989)  concluded  that  natural  selec- 
tion rather  than  lack  of  gene  flow  may  have  differentiated 
the  populations.  They  also  suggested  that  seed  dispersal 


47 


by  Clark's  nutcrackers  is  likely  to  be  "a  more  important 
component  of  gene  flow  in  limber  pine  than  pollen  trans- 
fer, at  least  across  elevational  gradients." 

Schuster  and  Mitton  (1992)  also  compared  five  popula- 
tions of  limber  pine  across  a  100-km  range  and  found  low 
levels  of  differentiation  (mean  Gg^  =  0.035).  Hamrick  (un- 
published data  cited  in  Betancourt  and  others  1991)  ex- 
amined eight  populations  of  pinyon  pine  (Pinus  edulis) 
and  found  a  slightly  greater  differentiation  (mean  Gg^  = 
0.077). 

Nine  populations  of  the  Siberian  stone  pine  were  stud- 
ied by  Krutovskii  £ind  others  (1989)  to  examine  interpopu- 
lation  genetic  differentiation.  Their  results  indicated  only 
slight  differences  among  populations.  Using  Nei's  meas- 
ures of  diversity,  they  determined  that  on  average  about 
98  percent  of  allozyme  variation  was  attributed  to  intra- 
population  variability  and  only  2  percent  to  interpopula- 
tion  variability  (mean  Gg^  =  0.022).  However,  they  did 
find  a  significant  correlation  between  geographical  dis- 
tance and  genetic  differences  among  populations,  but  only 
for  populations  separated  by  400  km  or  more.  They  also 
found  a  higher  level  of  differentiation  among  populations 
within  the  same  region  than  between  regions,  which  may 
reflect  adaptation  to  local  environmental  conditions.  The 
high  level  of  intrapopulational  variation  in  Siberian  stone 
pine  is  very  similar  to  what  has  been  found  for  most  wind- 
dispersed  conifer  species  (Krutovskii  and  others  1989). 

EVOLUTIONARY  IMPLICATIONS 

Tomback  and  Linhart  (1990)  speculated  that  bird- 
dispersed  pines  might  be  more  prone  to  genetic  drift  than 
are  wind-dispersed  species.  The  ranges  of  these  pines, 
particularly  subalpine  species,  tend  to  be  highly  frag- 
mented, and  small,  isolated  populations  are  not  uncom- 
mon. These  smaller  populations  may  show  founder  ef- 
fects, particularly  loss  of  rare  alleles.  In  addition,  because 
of  the  occurrence  of  tree  clusters  and  the  potential  for 
cross-pollination  among  cluster  members,  there  may  be 
high  rates  of  inbreeding  in  these  populations.  When  effec- 
tive population  size  is  smaller  than  100,  genetic  drift  can 
be  far  more  potent  in  causing  gene  frequency  changes 
than  natural  selection  (Futuyma  1986  and  references 
therein).  It  is  possible  that  the  evolution  of  cone,  seed, 
and  tree  morphology  in  bird-dispersed  pines  was  in  part 
the  consequence  of  new  traits  that  arose  from  a  combina- 
tion of  drift  and  inbreeding  in  small,  isolated  populations. 
Nutcrackers,  or  ancestral  corvids,  selected  for  preferred 
traits  and  spread  these  traits  to  other  populations  or 
founded  new  populations  with  these  traits. 

If  bird-dispersed  pines  have  this  population  structure, 
Tomback  and  Linhart  (1990)  make  two  predictions:  "(1) 
there  is  greater  genetic  variation  among  populations  of 
bird-dispersed  pines  than  among  wind-dispersed  pines, 
and  (2)  there  are  small,  isolated  populations  of  bird- 
dispersed  pines  with  highly  differentiated  gene  pools  com- 
pared to  one  another  and,  particularly,  compared  to  large 
populations  that  may  regularly  experience  some  gene  flow 
from  other  large  populations." 

Although  the  earliest  populations  of  bird-dispersed  pines 
may  have  had  the  characteristics  described  by  Tomback 


and  Linhart  (1990),  the  findings  of  Krutovskii  and  others 
(1989),  Schuster  and  Mitton  (1992),  and  Hamrick  (unpub- 
lished data  cited  in  Betancourt  and  others  1991)  do  not 
support  the  first  prediction,  and  it  may  well  be  that  the 
great  distances  over  which  nutcrackers  disseminate  seeds 
provide  effective  gene  flow.  Betancourt  and  others  (1991) 
describe  a  disjunct  population  of  Colorado  pinyon  pine  in 
northern  Colorado  that  was  the  result  of  a  founding  event 
probably  about  400  years  ago.  Compared  to  other  pinyon 
pine  populations,  the  average  percentage  of  polymorphic 
loci  and  nimiber  of  alleles  per  locus  were  reduced,  although 
levels  of  heterozygosity  were  not  affected.  These,  how- 
ever, are  typical  founder  effects  (Betancourt  and  others 
1991;  Futuyma  1986).  Within  the  last  century,  the  found- 
ing population  has  given  rise  to  four  descendant  popula- 
tions, three  with  fewer  than  50  trees  each  (Premoli  and 
others  1993).  The  two  smallest,  most  distant  descendant 
populations  had  fewer  alleles  per  polymorphic  locus  and 
reductions  in  individual  heterozygosity  compared  to  the 
larger  populations.  The  mean  Gg^  value  of  0.060  indicated 
little  differentiation  among  populations.  Because  tree 
clusters  do  not  occur  often  in  Colorado  pinyon  pine,  these 
new  populations  may  not  adequately  test  prediction  2. 
However,  they  do  demonstrate  that  seed  dispersal  by 
birds  results  in  small,  isolated  foimding  populations. 
Much  more  work  is  required  to  thoroughly  characterize 
the  genetic  population  structure  of  bird-dispersed  pines 
and  to  address  questions  concerning  the  biology  of  tree 
clusters. 

ACKNOWLEDGMENTS 

We  thank  Konstantin  Krutovskii  and  Dmitri  Politov  for 
providing  translations  of  their  papers  £ind  for  interesting 
discussions  during  the  workshop.  In  addition,  we  are 
grateful  to  Jeff  Karron,  Andrea  Premoli,  and  Kathleen 
Shea  for  helpful  comments  on  an  earlier  draft  of  the  pa- 
per. Support  for  DFT  for  attending  the  SSPE  Workshop 
was  provided  by  the  College  of  Liberal  Arts  and  Sciences, 
University  of  Colorado  at  Denver. 

REFERENCES 

Betancourt,  J.  L.;  Schuster,  W.  S.;  Mitton,  J.  B.; 
Anderson,  R.  S.  1991.  Fossil  and  genetic  history  of 
a  pinyon  pine  (Pinus  edulis)  isolate.  Ecology.  72(5): 
1685-1697. 

Bibikov,  D.  I.  1948.  On  the  ecology  of  the  nutcracker. 
Trudy  Pechorskogo-Ilychskogo  Gosudarstvennogo 
Zapovednika.  4(2):  89-112.  [In  Russian,  translated 
by  L.  Kelso]. 

Bullock,  S.  J.  1981.  Aggregations  of  Prunus  ilicifolia 
(Rosaceae)  during  dispersal  and  its  effect  on  survival 
and  growth.  Madrono.  28:  94-95. 

Carsey,  K.  S.;  Tomback,  D.  F.  1992.  Growth  form  distribu- 
tion and  genetic  relationships  in  tree  clusters  of  Pinus 
flexilis,  a  bird-dispersed  pine.  Manuscript. 

Critchfield,  W.  F.;  Little,  E.  L.,  Jr.  1966.  Geographic 
distribution  of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service.  97  p. 


48 


Epperson,  B.  K  1990.  Spatial  patterns  of  genetic  varia- 
tion within  plant  populations.  In:  Brown,  A.  H.  D.; 
Clegg,  M.  T.;  Kahler,  A.  L.;  Weir,  B.  S.,  eds.  Plant  popu- 
lation genetics,  breeding,  and  genetic  resources. 
Sunderland,  MA:  Sinauer  Associates:  229-253. 

Feldman,  R.  1991.  Growth  form  and  reproductive  output 
in  limber  pine:  the  cost  of  mutualism.  Denver,  CO:  Uni- 
versity of  Colorado  at  Denver,  Department  of  Biology. 
46  p.  Thesis. 

Feldman,  R.;  Tomback,  D.  F.  1991.  Growth  form  and  re- 
productive output  in  limber  pine:  the  cost  of  mutualism. 
Manuscript. 

Fumier,  G.  R.;  Knowles,  P.;  Clyde,  M.  A.;  Dancik,  B.  P. 
1987.  Effects  of  avian  seed  dispersal  on  the  genetic 
population  structure  of  whitebark  pine  populations. 
Evolution.  4(3):  607-612. 

Futuyma,  D.  J.  1986.  Evolutionary  biology.  2d  ed. 
Sunderland,  MA:  Sinauer  Associates.  600  p. 

Hamrick,  J.  L.;  Godt,  M.  J.  W.  1990.  Allozyme  diversity  in 
plant  species.  In:  Brown,  A.  H.  D.;  Clegg,  M.  T.;  Kahler, 
A.  L.;  Weir,  B.  S.,  eds.  Plant  population  genetics,  breed- 
ing, and  genetic  resoiirces.  Simderland,  MA:  Sinauer 
Associates:  43-63. 

Hayashida,  M.  1989.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecol- 
ogy and  Management.  28:  115-129. 

Holtmeier,  F.  -K.  1986.  Uber  Bamninseln  (Kollektive) 
an  der  klimatischen  Waldgrenze — imter  besonderer 
Beriicksichtigung  von  Beobachtungen  inverschiedenen 
Hochgebirgen  Nord-amerikas.  Wetter  und  Leben.  38: 
121-139. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Keeley,  J.  E.  1988.  Population  variation  in  root  grafting 
and  a  hypothesis.  Oikos.  52(3):  364-366. 

Kephart,  S.  R.  1990.  Starch  gel  electrophoresis  of  plant 
isozymes:  a  comparative  analysis  of  techniques.  Ameri- 
can Journal  of  Botany.  77(5):  693-712. 

Knowles,  P.  1984.  Genetic  v£iriability  among  and  within 
closely  spaced  populations  of  lodgepole  pine.  Canadian 
Journal  of  Genetics  and  Cytology.  26:  177-184. 

Krutovskii,  K.  V.;  Politov,  D.  V.;  Altukhov,  Y.  P.;  Milyutin, 
L.  I.;  Kuznetsova,  G.  V.;  Iroshnikov,  A.  I.;  Vorob'ev,  V.  N.; 
Vorob'eva,  N.  A.  1989.  Genetic  variability  in  the  Sibe- 
rian stone  pine,  Pinus  sibirica  Du  Tour.  IV.  Genetic 
diversity  and  the  extent  of  genetic  differentiation  be- 
tween populations.  Genetika.  25(11):  2009-2032.  [Trans- 
lation published  by  Pleniun  Publishing  Corp.  in  1991]. 

Laimer,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in 
the  ecology  and  evolution  of  limber  and  whitebark 
pines.  In:  Dancik,  B.  P.;  Higginbotham,  K.  0.,  eds.  Sixth 
North  Americsin  forest  biology  workshop:  proceedings. 
Edmonton,  AB:  University  of  Alberta:  15-48. 

Lepper,  M.  G.  1974.  Pinus  flexilis  James  and  its  environ- 
mental relationships.  Davis,  CA:  University  of  California 
at  Davis,  Graduate  Group  in  Ecology.  197  p.  Dissertation. 

Ligon,  J.  D.  1978.  Reproductive  interdependence  of 
pinon  jays  and  pinon  pines.  Ecological  Monographs. 
48:  111-126. 


Linhart,  Y.  B.  1989.  Interactions  between  genetic  and  eco- 
logical patchiness  in  forest  trees  and  their  dependent 
species.  In:  Bock,  J.  E.;  Linhart,  Y.  B.,  eds.  Evolutionary 
ecology  of  plants.  Boulder,  CO:  Westview  Press:  1-31. 

Linhart,  Y.  B.;  Mitton,  J.  B.;  Sturgeon,  K.  B.;  Davis,  M.  L. 
1981.  Genetic  variation  in  space  and  time  in  a  popula- 
tion of  ponderosa  pine.  Heredity.  46:  407-426. 

Linhart,  Y.  B.;  Tomback,  D.  F.  1985.  Seed  dispersal  by 
nutcrackers  causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

Mattes,  H.  1982.  Die  Lebengemeinschaft  von  Tannenhaher 
imd  Arve.  Berichte  Eidenossische  Anstalt  fur  das 
forstliche  Versuchswesen,  Birmensdorf.  Nr.  241.  74  p. 

McCaughey,  W.  W.;  Schmidt,  W.  C;  Shearer,  R.  C.  1986. 
Seed-dispersal  characteristics  of  conifers  in  the  Inland 
Mountain  West.  In:  Shearer,  R.  C,  comp.  Proceedings — 
conifer  tree  seed  in  the  Inland  Mountain  West  sympo- 
sium; 1985  August  5-6;  Missoula,  MT.  Gen.  Tech.  Rep. 
INT-203.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Research  Station:  50-62. 

McNaughton,  G.  M.  1984.  Comparative  water  relations 
of  Pinus  flexilis  at  high  and  low  elevations  in  the  cen- 
tral Rocky  Mountains.  Laramie,  WY:  University  of 
Wyoming,  Department  of  Botany.  59  p.  Thesis. 

Mezhenny,  A.  A.  1961.  Food  competitors,  enemies  and  dis- 
eases. In:  Egorov,  O.  V.,  ed.  Ecology  and  economics  of 
the  Yakut  squirrel.  Moscow:  Akademiya  Nauk:  124-129. 

Politov,  D.  v.;  Krutovskii,  K.  V.  1990.  Genetic  variation  of 
Siberian  stone  pine  Pinus  sibirica  Du  Tom*.  V.  Analysis 
of  mating  system.  Genetika.  26(11):  2002-2011.  [Trans- 
lation published  by  Plenimi  Publishing  Corp.  in  1991]. 

Premoh,  A.  C;  Chischilly,  S.;  Mitton,  J.  B.  1993.  Levels 
of  genetic  variation  captured  by  four  descendant 
populations  of  pinyon  pine  (Pinus  edulis  Engelm.). 
Manuscript. 

Quellar,  D.  C;  Goodnight,  K.  F.  1989.  Estimating  relate d- 
ness  using  genetic  markers.  Evolution.  43:  258-275. 

Schuster,  W.  S.  F.;  Alles,  D.  L.;  Mitton,  J.  B.  1989.  Gene 
flow  in  limber  pine:  evidence  from  pollination  phenology 
and  genetic  differentiation  along  an  elevational  transect. 
American  Journal  of  Botany.  76(9):  1395-1403. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1991.  Relatedness  within 
clusters  of  a  bird-dispersed  pine  and  the  potential  for 
kin  interactions.  Heredity.  67:  41-48. 

Schuster,  W.  S.  F.;  Mitton,  J.  B.  1992.  Gene  flow  in 
limber  pine.  II.  Data  for  five  populations.  Manuscript 
in  preparation. 

Swanberg,  P.  O.  1956.  Territory  in  the  thick-billed  nut- 
cracker A^uci/ra^a  caryocatactes.  Ibis.  98:  412-419. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. The  Living  Bird.  16:  123-161. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1988.  Nutcracker-pine  mutualisms:  multi- 
tnmk  trees  and  seed  size.  In:  Ouellet,  H.,  ed.  Acta  XIX 
Congressus  Internationalis  Ornithologici,  Vol.  1;  1986 
June  22-29.  Ottawa,  ON:  University  of  Ottawa  Press: 
518-527. 


49 


Tomback,  D.  F.  1988.  [Unpublished  observations  in  the 
Upper  Engadine  Valley  of  Switzerland,  October  1988]. 

Tomback,  D.  F.;  Knowles,  J.  W.  1989.  Post-fire  whitebark 
pine  seed  dispersal  by  Clark's  nutcracker  in  Yellowstone 
National  Park.  Unpublished  data  and  observations. 

Tomback,  D.  F.;  Linhart,  Y.  B.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution  of  whitebark  pine  and  nutcrackers:  implications 
for  forest  regeneration.  In:  Schmidt,  W.  C;  McDonald, 
K  J.,  comps.  Proceedings — sjnnposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  118-129. 

Tomback,  D.  F.;  Holtmeier,  F.  -K.;  Mattes,  H.;  Carsey, 
K.  S.;  Powell,  M.  L.  1993.  Tree  clusters  and  growth  form 


distribution  in  Pinus  cembra,  a  bird-dispersed  pine. 

Arctic  and  Alpine  Research.  [In  press]. 
Turcek,  F.  J.  1966.  Uber  das  Wiederauffinden  von  im 

Boden  versteckten  Samen  durch  Tannen-und 

Eichelhaher.  Waldhygiene.  6:  215-217. 
Vander  Wall,  S.  B.  1992.  The  role  of  animals  in  dispersing 

a  "wind-dispersed"  pine.  Ecology.  73:  614-621. 
Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations 

of  Clark's  nutcracker  and  the  pinon  pine  for  efficient 

seed  harvest  and  dispersal.  Ecological  Monographs. 

47:  89-111. 

Weaver,  T.;  Jacobs,  J.  1990.  Occvirence  of  multiple  stems  in 
whitebark  pine.  In:  Schmidt,  W.  C;  McDonald,  K  J., 
comps.  Proceedings — symposium  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
156-159. 


50 


Basic  Ecology 


International  Workshop 
St.  Moritz  1 992 


EXPERIENCES  WITH  REPRODUCTION 
OF  CEMBRA  PINE 


Ernst  Frehner 
Walter  Schonenberger 


Abstract — ^As  regards  reproduction,  cembra  pine  {Pinus  cembra 
L.)  is  a  special  case.  If  the  greater  part  of  the  seeds  are  stolen  by 
nutcrackers,  they  should  be  harvested  before  they  are  ripe  and 
then  matured.  With  correct  treatment  it  is  possible  to  store 
seeds  for  up  to  7  years.  After  soaking  in  water,  germination  be- 
gan almost  at  once.  The  separate  steps  along  the  long  process 
from  harvesting  via  seed  extraction,  maturation,  storage,  testing, 
sowing,  and  transplantation  to  planting  out  are  described  in 
detail. 


in  the  Alps,  and  that  the  existing  stands  would  diminish 
considerably  within  a  few  generations. 

Depending  on  the  number  of  cones  produced,  the  nut- 
cracker stores  seeds  as  winter  provisions  in  nimierous 
caches,  usually  numbering  over  10,000,  each  containing 
from  three  to  five  seeds,  but  only  consimies  about  80  per- 
cent of  its  reserve.  The  remaining  seeds  contribute 
greatly  to  the  regeneration  of  cembra  pine  stands.  The 
nutcracker  may  transport  seeds  over  a  distance  of  15  km 
or  more  and  over  a  difference  in  height  of  up  to  600  m. 


This  study  reports  our  experiences  over  many  years  of 
seed  £ind  afforestation  trials  with  cembra  pine  {Pinus 
cembra  L.),  from  seed  harvesting  to  planting. 

Cembra  pine  is  a  special  case  as  regards  reproduction 
(Frehner  and  Fuerst  1992).  The  extreme  conditions  in  its 
natural  environment  at  the  upper  timberline  strongly  in- 
fluence flowering,  fertilization,  and  seed  maturation,  so 
that  the  quantity  and  quality  of  seed  varies  from  year  to 
year  (table  1).  The  results  of  repeated  studies  on  cone 
quality,  maturation,  and  storage  vary  correspondingly. 

The  special  features  of  reproduction  in  cembra  pine  be- 
gin with  seed  collection.  Except  in  full  mast  years,  most 
of  the  seeds  have  been  eaten  or  collected  by  the  nut- 
cracker by  the  time  of  their  natural  maturation.  Like  the 
seeds  of  many  other  tree  and  shrub  species,  those  of 
cembra  pine  require  several  months  of  dormancy.  Matu- 
ration of  the  seed  and  time  of  extraction  is  therefore  very 
important  for  the  development  of  the  embryo.  To  better 
exploit  fiill  mast  years,  the  durability  of  the  seeds  should 
be  extended  through  improved  storage  methods.  It  is  also 
important  that  stored  seeds  should  be  induced  to  germi- 
nate and  produce  successful  saplings  at  short  notice.  Also 
in  terms  of  sowing,  transplantation,  and  potting,  cembra 
pine  requires  special  conditions. 

SYMBIOSIS  WITH  THE 
NUTCRACKER 

The  importance  of  the  nutcracker  for  the  natural  distri- 
bution of  cembra  pine  is  undisputed.  The  detailed  study 
by  Mattes  (1982)  on  this  relationship  concludes  that  with- 
out the  nutcracker  there  would  be  no  cembra  pine  forests 


CONE  COLLECTION 

In  years  with  normal,  moderate  fructification,  the  nut- 
cracker begins  to  pick  the  seeds  at  medium  altitudes  as 
soon  as  ripening  has  begun,  about  mid-August,  and  con- 
tinues to  do  so  until  about  the  end  of  September.  This 
means  that  collecting  cones  is  a  race  with  the  bird.  If 
cones  are  to  be  collected  in  years  with  partial  mast,  this 
must  be  done  at  an  early  stage,  while  they  are  still  un- 
ripe, before  the  birds  begin  their  robbery,  and  then  stored 
under  optimvmi  conditions  for  maturation. 

In  long-term  observations,  19  percent  of  the  years  pro- 
duced full  mast,  55  percent  partial  mast,  and  26  percent 
failed  to  produce  (Mattes  1982).  As  a  rule  it  is  only  in  fiill 
mast  years  that  enough  seeds  ripen  on  the  tree  to  meike 
harvesting  worthwhile;  in  this  case  the  nutcracker  does 
not  have  time  to  devour  or  carry  away  all  the  remaining 
seeds. 

Neither  color  nor  size  of  cones  has  any  influence  on  seed 
quality  or  planting  success.  The  costs  of  harvesting  and 
seed  extraction  are  considerably  greater  for  small  cones. 

With  early  harvesting  it  is  very  important  that  the 
seeds  should  be  allowed  to  ripen  imder  optimum  condi- 
tions of  moisture  and  warmth,  as  the  embryos  are  in  a 
very  early  stage  of  development  (only  up  to  3  mm  long). 

Studies  on  protected  cones  attached  to  mother  trees  at 
2,000  m  have  shown  that  the  length  of  the  embryo  hardly 
changes  between  mid-September  and  the  end  of  October. 
In  contrast,  embryos  in  cones  harvested  early  (mid- 
September)  have  elongated  by  3-4  mm.  This  shows  that 


Table  1 — Characteristics  of  cembra  pine  seed  crops 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11, 1992. 

Ernst  Frehner  is  former  Head  of  the  Research  Nursery,  Walter 
Schonenberger  is  Head  of  the  Research  Section  Silviculture,  Swiss 
Federal  Institute  for  Forest,  Snow  and  Landscape  Research,  CH-8903 
Birmensdorf,  Switzerland. 


Harvesting  time  per  tree 

Yield  per  tree 

Number  of  cones  per  tree 

Weight  per  cone 

Seed  yield  (fresh  weight) 

Number  of  living  seeds  per  cone 

Thousand-seed  weight 


about  1  hour 
about  15  kg  cones 
about  150-200 
60-100  g 
20-30  percent 
50-70 
300-400  g 


52 


Table  2 — Cone  collection,  maturation,  and  seed  extraction 

September  Cone  harvesting  (embryo  length  up  to  3  mm) 

September  Stratification  of  cones  for  maturation  at  1 0-20  °C 

for  8-1 0  weeks 

November  Seed  extraction  (embryo  length  3-5  mm) 


it  is  not  worthwhile  enclosing  whole  trees  in  nets  or  treat- 
ing the  cones  with  chemicals.  Early  harvesting  and  favor- 
able conditions  of  stratification  are  preferable  to  late 
harvesting. 

CONE  MATURATION 

Stratification  (storing  cones  in  layers  in  a  moist,  airy 
substrate)  is  the  simplest  and  most  certain  technique  of 
accelerating  ripening  and  reducing  inhibition  of  germi- 
nation so  that  the  seeds  are  ready  to  germinate  quickly. 
Stratification  simulates  the  natural  conditions  of  moisture 
and  warmth  prevailing  in  the  forest,  or  may  even  be 
better. 

The  fi-eshly  harvested  cones  are  stored  in  layers  at  a 
proportion  of  one  part  cones  to  five  parts  substrate.  Suit- 
able substrates  are  washed  sand  or  a  mixture  of  sand  and 
peat.  The  substrate  must  be  sieved  so  that  its  particles 
are  smaller  than  the  seeds.  The  stratified  cones  should 
then  be  stored  in  a  cupboardlike  container  with  good 
drainage  that  protects  the  cones  from  attack  by  mice  or 
birds. 

The  development  of  the  embryos  depends  greatly  on  the 
temperature  of  the  substrate  and  the  water  content  of  the 
cones.  Studies  have  shown  that  the  embryos  develop  only 
slowly  at  low  temperatures,  while  high  temperatures  also 
have  a  negative  influence.  The  best  results  were  obtained 
with  substrate  temperatures  alternating  between  10  and 
20  °C.  It  is  advisable  to  keep  the  stratified  cones  in  a  pro- 
tected place  (for  example,  an  unheated  plastic  tiinnel  or 
glasshouse),  where  the  temperature  may  rise  toward 
25  °C  during  the  day  and  sink  to  about  5  °C  at  night. 
Naturally,  the  temperatiire  within  the  substrate  does  not 
vary  so  greatly. 

Moisture  is  very  important.  The  substrate  should  be 
moist  and  airy,  and  the  cones  should  be  just  moist  enough 
to  prevent  the  seed  coat  from  shrinking  away  from  the 
endosperm.  Maturation  to  the  point  of  seed  extraction 
proceeds  much  better  under  moist  conditions  than  with 
dry  storage.  With  ripening  under  dry  conditions  the  ger- 
mination rate  is  50  percent  lower  than  that  after  moist 
storage,  while  dormancy  is  10  times  longer. 

SEED  EXTRACTION 

After  8  to  10  weeks'  stratification,  the  seeds  can  be  ex- 
tracted. The  cones,  which  are  by  this  time  beginning  to 
disintegrate,  are  sieved  out  of  the  substrate,  left  to  dry  for 
a  while,  and  then  rubbed  through  finer  sieves  to  separate 
the  seeds  from  scales  and  cone  axes.  The  finer  scales  and 
substrate  particles  can  then  be  removed  through  further 


sieving  and  washing  in  a  water  bath,  either  immediately 
or  after  another  ripening  process.  In  a  water  bath  the  im- 
fertilized  seeds  (empty  seed  cases)  float  to  the  top  and  can 
be  decanted.  To  summarize,  the  course  of  cone  matura- 
tion is  shown  in  table  2. 

SEED  MATURATION 

The  treatment  of  the  seeds  after  extraction  depends  on 
their  intended  use  and  how  long  they  are  to  be  stored. 

Under  natural  conditions  it  takes  2  to  3  years  for  the 
seeds  to  germinate  after  the  cones  have  fallen.  This  may 
be  due  to  many  factors,  for  example,  hard  and  imperme- 
able seed  cases,  incomplete  development  of  the  embryo,  or 
inhibiting  substances  in  the  endosperm,  in  the  seed  case, 
or  elsewhere  in  the  seed.  In  cembran  pine,  the  main  rea- 
son is  the  low  stage  of  development  of  the  embryos:  at  the 
time  of  harvesting  the  embryos  are  not  more  than  3  mm 
long. 

With  artificial  maturation  using  variable  temperatures, 
the  period  between  harvesting  and  germination  can  be  re- 
duced to  one  winter.  Here,  the  aim  is  to  induce  the  em- 
bryos to  grow  fi"om  a  length  of  3  mm  or  less  to  the  germi- 
nation length  of  7-9  mm  (table  3).  To  promote  embryo 
growth  even  after  seed  extraction,  the  seeds  are  alter- 
nately exposed  to  warmth  and  cold  in  a  monthly  rotation. 
Exposure  to  warmth  for  periods  of  more  than  4  to  6  weeks 
does  not  positively  influence  embryo  development.  Great 
attention  should  be  paid  to  water  content  throughout  the 
whole  period  of  seed  treatment,  as  during  cone  storage. 
The  better  the  seed  maturation,  the  better  the  embryo  de- 
velopment and,  consequently,  the  better  the  seeds  are 
suited  for  drying  and  storage. 

BASIC  PRINCIPLES  OF  STORAGE 

After  drying,  the  seeds  are  placed  in  one  of  the  con- 
tainer types  mentioned  and  stored  imder  cool  conditions. 
In  the  early  stages  they  should  be  regularly  inspected. 

Water  content: 

•  the  lower  the  water  content,  the  lower  the  storage 
temperature; 

•  the  lower  the  storage  temperature,  the  longer  the 
seeds  can  be  stored; 

•  the  longer  the  seeds  have  been  stored,  the  longer 
they  must  be  stratified  before  sowing. 


Table  3 — Results  of  the  maturation  of  cembran  pine  seeds 
(harvested  September  22)  after  warm-cold-warm 
treatment 


Features 

September  23 

February  28 

Seed  length  (peeled) 

9.5  mm 

10.4  mm 

Channel  length 

4.5  mm 

8.5  mm 

Embryo  length 

1.0  mm 

8.4  mm 

Seeds  with  visible  embryos 

11  percent 

72  percent 

Seeds  ready  to  germinate 

96  percent 

76  percent 

53 


Storage  containers: 

•  at  high  water  content  (>20  percent),  in  perforated 
plastic  bags; 

•  at  low  water  content  (<15  percent),  in  airtight  con- 
tainers accommodating  2-5  kg. 

Storage  rooms:  climatized: 

•  for  short-term  storage  with  high  water  content, 
1-3  °C; 

•  for  fairly  long  storage  with  low  water  content, 
-5  to  -10  °C. 

Freeze  drying  has  proved  unsuitable  for  cembran  pine 
seeds,  because  the  water  content  is  too  high. 

Fungal  attack:  for  prevention,  the  seeds  should  be 
treated  with  a  fungicide  and  regularly  examined  as  to 
smell  and  condensed  moisture  indicating  a  too-high  water 
content  in  the  containers;  if  this  occurs,  the  seeds  should 
be  dried  immediately. 

STORAGE  OF  MATURE  SEEDS 

Without  treatment  for  reduction  of  water  content  the 
seeds  cannot  be  stored  longer  than  one  winter.  With  re- 
duction of  water  content  to  about  25  percent,  the  seeds 
can  be  stored  at  1-3  °C  for  up  to  3  years  without  great  loss 
in  the  germination  rate. 

Later  trials  showed  that  gently  handled  seeds  could  be 
dried  to  a  water  content  of  15  percent,  and  that  under 
storage  at  -5  to  -10  °C  the  germination  capacity  did  not 
decrease  for  up  to  7  years.  A  water  content  of  more  than 
20  percent  rendered  storage  below  0  °C  iinfeasible. 

THREE  SEED  TREATMENTS 

After  seed  extraction  three  treatments  for  three  differ- 
ent storage  periods  can  be  defined  as: 

•  Treatment  1:  for  immediate  sowing  without  storage 
(table  4). 

•  Treatment  2:  for  storage  over  2  to  3  years  (table  5). 

•  Treatment  3:  for  storage  over  3  to  7  years  (table  6). 

GERMINATION  TRIALS 

Germination  trials  begin  in  December.  The  seeds  are 
sown  in  a  sterile  peat/sand  in  sterilized  containers,  being 

Table  4 — Treatment  1 :  for  immediate  sowing  without  storage 

November  Seed  stratification  outdoors  for  5-6  months; 

substrate  three  parts  washed  sand  to  one  part 
seeds 

April  Sowing  (usually  30-50  percent  of  seeds  lie-over) 

Advantages:        No  need  for  climatization 

Disadvantages:    No  storage  possible,  annual  seed  harvesting 

necessary,  high  proportion  of  lie-over,  high  risk 
of  robbery  by  birds  and  mice 


Table  5 — Treatment  2:  for  storage  over  2-3  years  with  high  water 

content  under  refrigeration  in  plastic  bags  (for  example,  in 
years  with  partial  or  full  mast) 


November  Seed  stratification  for  4-6  weeks  at  1 6-1 8  °C  in 

washed  sand 

December/January  Sieving  and  cleaning  of  seeds;  drying  of  seeds 
for  1  -2  weeks  at  1 6-1 8  °C  to  about 
25  percent  water  content:  embryo  length 
increases  to  5-7  mm 

January  Storage  of  seeds  in  perforated  plastic  bags  at 

1  -3  °C  over  2-3  years 

October  Seeds  are  soaked  for  1  -2  days;  stratification 

outdoors  for  6  months  in  washed  sand 

April/May  Sowing  (no  lie-over  to  be  expected) 


Advantages:  Allows  storage  for  up  to  3  years  without  lie-over 

Disadvantages:       Storage  climatized  at  1-3  °C  necessary,  high 
water  content,  risk  of  rotting  and  fungal 
infection,  long  stratification  period 


normally  spread  and  pressed  down,  then  covered  with 
about  1  cm  sand  or  substrate.  The  trials  are  best  con- 
ducted in  rooms  without  direct  sunlight;  the  ideal  tem- 
perature is  16  to  18  °C.  Alternating  temperatures  have 
a  positive  effect.  The  seeds  should  be  watered  by  hand 
according  to  need.  Storage  in  greenhouses  with  automatic 
spraying  has  not  proved  suitable.  The  germination  tests 
take  6-8  weeks. 


Table  6 — Treatment  3:  for  storage  in  airtight  containers  for  3-7  years 
with  low  water  content  and  freezing  temperatures  (for 
example,  for  seeds  harvested  In  good  partial  or  full  mast 
years) 


December 

Seed  stratification;  first  warm  treatment  at 

16-18  °C  for  4-5  weeks 

January 

Cold  treatment  at  1  -3  °C  for  4-5  weeks 

February 

Second  warm  treatment  at  1 6-1 8  °C  for 

4-5  weeks  until  germination  begins;  embryo 

length  increases  to  7-9  mm 

February 

Sieving  and  cleaning  of  seeds;  drying  at 

1 6-1 8  °C  for  1  -2  weeks  tol  5  percent  water 

content 

March 

Storage  in  airtight  plastic  bags  or  glass 

containers  at  -5  to  -1 0  °C  for  3-7  years 

April 

Defrosting  for  1-2  days,  soaking  in  lukewarm 

water  for  1-2  days,  sowing 

Advantages: 

Seeds  can  be  stored  for  up  to  7  years,  no 

lie-over,  short-term  stratification  (6-8  weeks). 

good  mast  years  and  seed  quality  can  be 

fully  exploited 

Disadvantages: 

Long-term  deep-freezing  required 

54 


SOWING 

The  seeds  are  sown  between  the  end  of  April  and  mid- 
May.  The  number  of  seeds  to  be  sown  depends  on  the  re- 
sults of  the  germination  trials.  Because  germination  is 
slow,  the  seeds  should  be  sown  in  boxes  resistant  to  mice 
and  birds.  The  best  substrates  are  soils  without  lime- 
stone mixed,  according  to  conditions  and  pH,  with  peat 
(which  loosens  the  soil,  maintains  a  pH  between  5  and  6, 
and  regulates  the  moisture  level)  or  raw  humus  contain- 
ing mycorrhizal  fungi  from  cembran  pine  stands.  The 
seed  and  transplantation  beds  should  continuously  be 
occupied  by  cembran  pines  except  for  short  periods  with 
green  fertilizing. 

Cembran  pine  grows  best  in  thickly  sown  beds,  with 
1,200  to  1,800  seedlings  per  square  meter,  grown  for  2 
to  3  years  in  a  seedbed. 

If  the  seeds  begin  to  germinate  during  stratification, 
they  should  be  sown  immediately.  If  this  is  unfeasible, 
they  can  be  frozen  at  -3  to  -4  °C  and  sown  as  required. 
Seeds  that  have  already  germinated  and  whose  radicle  is 
longer  than  1  cm  should  not  be  planted  in  a  seedbed,  but 
thinned  out;  otherwise  the  resulting  plants  are  liable  to 
be  worthless,  with  tortuous  root  collars,  short  stems, 
small  crowns,  and  few  fine  roots.  Size,  quality,  and  yield 
are  definitely  better  in  thinned  seedlings.  Shortening  the 
roots  causes  deformations. 

We  did  not  run  any  trials  with  sowing  in  autumn,  as 
the  risks  of  attack  by  mice  and  birds  and  also  snow  pres- 
sure on  the  protection  lattice  are  then  very  high. 

TRANSPLANTING 

The  specially  designed  beds  are  continuously  supple- 
mented with  acid  soil  (for  example,  peat)  and  if  necessary 
with  chalk-free  acid  fertilizer.  To  hinder  invasion  by 
weeds  and  frost  heaving  and  to  restrict  evaporation,  and 


also  to  maintain  a  good  soil  structiire,  the  beds  are 
mulched. 

It  is  recommended  that  3-year-old  saplings  be  trans- 
planted in  spring  or  summer  (from  mid-July).  This  can  be 
done  by  machine,  with  a  spacing  of  10  by  18  cm.  Saplings 
with  poor  root  development  should  be  discarded,  as  it  is 
not  worth  planting  them  out.  The  saplings  stay  in  the 
transplanting  bed  for  3  years. 

POTTING 

High-altitude  afforestations  are  exposed  to  high  risks 
from  extreme  weather  conditions,  snow,  and  pests.  Expe- 
rience has  shown  that  seedlings  in  peat-fiber  pots  succeed 
better  than  those  with  naked  roots,  especially  on  diffic\ilt 
sites,  so  potting  is  recommended.  The  production  of  pot- 
ted plants  has  been  described  by  Schonenberger  and 
others  (1990).  It  is  important  to  keep  the  pots  well 
watered.  The  saplings  should  only  be  planted  out  when 
the  root  tips  begin  to  penetrate  through  the  peat-fiber 
pots. 

REFERENCES 

Frehner,  Ernst;  Fiirst,  Ernst.  1992.  Vom  Samen  bis  zur 
Pflanze.  Ein  Erfahrungsbericht  aus  dem  Forstgarten. 
Ber.  Eidgenoss.  Forsch.  anst.  Wald  Schnee  Landsch. 
333.  47  p. 

Mattes,  Hermann.  1982.  Die  Lebensgemeinschaft  von 
Tannenhaher  und  Arve.  Eidg.  Anst.  forstl. 
Versuchswes.,  Ber.  241,  74  S. 

Schonenberger,  Walter;  Frey,  Werner;  Leuenberger, 
Franz.  1990.  Oekologie  und  Technik  der  Aufforstung 
im  Gebirge  -  Anregungen  fiir  die  Praxis.  Eidg.  Anst. 
forstl.  Versuchswes.,  Ber.  325,  58  S. 

Translation:  Margaret  Sieber 


55 


DISTRIBUTION  AND  ECOLOGY  OF 
SIBERIAN  STONE  PINE  IN  THE  URALS 


P.  L.  Gorchakovsky 


Abstract — The  northern  and  southern  boundaries  of  the  distri- 
bution area  of  Siberian  stone  pine  (Pinus  sibirica)  in  the  Urals 
and  in  the  adjacent  plains  are  examined;  the  main  sociological 
factors  limiting  boundary  positions  are  described.  There  is  a 
natural  tendency  of  this  species  to  spread  northward  and  south- 
ward, from  places  recently  invaded,  but  this  process  is  usually 
stopped  or  turned  back  by  human  influence.  Numerous  occur- 
rences of  Siberian  stone  pine  beyond  its  present  continuous  range 
indicate  that  the  former  distribution  area  of  this  species  was 
larger  in  the  past  than  at  present. 


Siberian  stone  pine  (Pinus  sibirica),  or  briefly  Siberian 
pine,  is  considered  to  be  one  of  the  most  valuable  species 
of  coniferous  trees  growing  in  Russia.  Research  on  the 
limits  of  natural  distribution  of  Siberian  pine,  on  their 
causes  and  dynamics,  is  of  indisputable  theoretical  and 
practical  interest.  Research  is  especially  necessary  for 
scientific  substantiation  of  rational  use  of  all  forests,  par- 
ticularly those  with  Siberian  pine,  for  the  definition  of 
measures  to  protect  it  fi"om  being  exterminated,  and  for 
solving  the  problems  concerning  possible  cultivation  of 
this  species  in  adjacent  regions. 

The  general  distribution  area  of  Siberian  pine  includes: 
(1)  northeastern  regions  of  the  European  part  of  Russia 
(the  farthest  west  locality  is  in  the  upper  reaches  of  the 
Vychegda  River);  (2)  the  Ural  Mountain  range;  (3)  west- 
em  Siberia;  (4)  middle  and  eastern  Siberia  (up  to  the 
Aldan  River);  and  (5)  mountain  ranges  of  Altai,  the  Trans- 
baikal  area,  and  northern  Mongolia. 

The  main  limiting  factors  of  the  distribution  of  this 
species  are:  heat  deficiency  in  the  north;  permafi*ost,  peat- 
Ismd,  and  severe  winter  in  the  northeast;  water  deficiency 
and  low  air  hvunidity  in  the  south. 

The  western  part  of  the  area  is  located  in  the  Ural 
Mountains  and  in  the  northeastern  regions  of  the  Euro- 
pean part  of  Russia  adjacent  to  them.  Siberian  pine  occu- 
pies rather  large  areas  within  the  territory  of  Sverdlovsk 
and  Perm  provinces,  Komi  Republic,  and  within  Knanty- 
Mansy  national  district  of  Tyimien  province. 

THE  NORTHERN  LIMIT 

The  northern  border  of  the  distribution  area  (fig.  1) 
passes  along  the  Izhma-Pechora  watershed,  advances 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

P.  L.  Gorchakovsky  is  Professor,  Doctor  of  Biology,  Chief  Scientist  at 
the  Institute  of  Plant  and  Animal  Ecology,  Russian  Academy  of  Sciences, 
Ural  Branch,  620219,  Ekaterinburg,  Russia. 


considerably  northward  (up  to  65°30'  N.  latitude)  along 
the  valley  of  the  Pechora  River  and  even  farther  north- 
ward (67°  N.  latitude — the  northern  polar  circle).  Then 
it  switches  back  south  in  the  form  of  a  tongue,  to  the  foot- 
hills of  the  elevated  eastern  part  of  the  Russian  plain. 
From  Telpos-Iz  Mountain,  64°  N.  latitude,  it  leaps  north- 
ward along  the  eastern  slope  of  the  Poleir  and  Prepolar 
Urals  reaching  66°  10'  N.  latitude.  It  then  recedes  south- 
ward in  the  space  between  the  rivers,  but  returns  north- 
ward again  in  a  tongue-shaped  projection  along  the  vsdley 
of  the  Ob  River. 

Besides  that,  a  few  isolated  occurrences  of  Siberian  pine 
are  known  outside  its  main  distribution  area. 

The  northern  limit  of  Siberian  pine  does  not  reach  the 
extreme  boundaries  of  the  distribution  of  larch  (Larix 
sibirica)  and  spruce  (Picea  obovata).  The  well-known 
Russian  botanist  B.  N.  Gorodkov  (1929)  supposed  that  the 
advancement  of  Siberisin  pine  to  the  north  is  restricted  be- 
cause its  seeds  do  not  get  ripe  under  the  severe  conditions 
of  the  short  growing  season. 

However,  the  latest  observations  demonstrate  that 
even  at  the  northern  limit  of  the  distribution  area  seeds 
of  Siberian  pine  usually  get  ripe.  The  main  reason  likely 
is  that  Siberian  pine  is  not  able  to  develop  its  root  system 
imder  low  soil  temperature  (Tyrtikov  1954).  In  cold  soils, 
the  roots  of  Siberian  pine  ramificate  comparatively  weakly 
and  grow  more  slowly  than  the  roots  of  spruce  and  larch. 
Consequently,  insufficient  soil  heating  during  the  growing 
season  impedes  normal  development  of  root  systems,  and 
this  appears  to  be  the  main  factor  limiting  the  distribu- 
tion of  Siberian  pine  in  the  north.  Under  the  severe  cli- 
matic conditions  of  northern  Siberia,  pine  does  not  grow 
on  cold,  peatland  soils. 

Close  to  the  northern  boundaries,  in  the  upper  reaches 
of  the  Pechora  River,  Siberian  pine  selects  the  warmest 
well-heated  locations,  inhabiting  even  the  riverside  lime- 
stone outcrops  and  cliffs.  Contrary  to  that,  in  the  regions 
farther  south,  for  example  in  the  middle  and  southern 
Urals,  Siberian  pine  never  grows  in  such  habitats  because 
the  riverside  limestone  cliffs  are  too  warm  during  daylight 
periods. 

In  the  plains  adjacent  to  the  Urals  (the  Pechora  and 
western  Siberian  lowlands),  the  far-north  occurrences  of 
Siberian  pine  are  restricted  to  the  valleys  of  the  large  riv- 
ers (the  Pechora,  the  Ob,  and  some  of  their  tributaries). 
The  bovmdary  deviates  northward  in  the  form  of  narrow 
strips  along  the  river  valleys.  In  marshy  areas  between 
the  rivers,  however,  where  superficial  bedding  of  perma- 
fi'ost  fi*equently  occurs,  the  border  line  is  farther  south. 
Under  plain  conditions,  the  river  valleys  characteristically 
have  more  permeable  ground,  lower  levels  of  permafi"ost 
bedding,  and  warmer  soils  that  are  more  favorable  to 


56 


Figure  1 — Northern  distribution  limit  of  Siberian  stone  pine  in  tine  Urals  and  in  the  adjacent 
plains.  *  =  isolated  occurrences. 


Siberian  pine.  Moreover,  one  should  bear  in  mind  that 
the  valleys  of  the  large  rivers  that  bring  masses  of 
comparatively  warm  water  from  the  south  are  character- 
ized by  a  warmer  local  climate  favorable  to  the  growth  of 
Siberian  pine. 

At  the  Lyapin  depression,  in  the  region  of  Saranpaul, 
Siberian  pine- dominated  forests  are  to  be  found  on  buried 
palsa  peatbogs  covered  by  sandy  loam  deposits.  On  hill- 
ocks, better  conditions  for  drainage  and  heating  of  the  su- 
perficial layer  of  accumulated  deposits  favor  the  formation 
of  rather  tall  Siberian  pine  forests.  On  the  other  hand, 
raised  bogs  with  pools  and  ridges  between  the  hillocks  are 
treeless.  The  level  of  permafrost  here  is  rather  high;  ice- 
lenses  are  foimd  inside  the  peaty  hillocks. 

In  the  plains,  at  the  extreme  Hmit  of  its  distribution 
area,  Siberian  pine  is  usually  represented  by  solitary 
specimens  dispersed  within  spruce-larch-birch  forests; 
trees  usually  are  healthy,  without  signs  of  atrophy;  their 
tnmks  reach  10  to  14  m  in  height. 

Within  the  western  and  eastern  outskirts  of  the  Ural 
Moimtain  range,  the  Siberian  pine  border  displaces  north- 
ward along  meridionally  oriented  low  foothills  and  moim- 
tain ridges.  There,  this  species  grows  on  slopes  or  near 
brooks  and  small  rivulets  imder  moderate  or  slightly  in- 
creased (flowing)  ground  moisture  conditions.  The  advance 
northward  is  promoted  by  a  variety  of  sites  suitable  for  for- 
est growth  as  well  as  by  favorable  drainage,  better  soil 
heating,  and  consequently  by  lower  depth  of  permafrost. 

In  the  watershed  of  the  Urals  there  are  huge  mountains 
topped  by  woodless  tundrahke  "goltsy"  summits,  located 


within  the  mountain  tundra  and  high-mountain  cold 
desert  belts  (Gorchakovsky  1975,  1989).  They  are  char- 
acterized by  a  more  severe  climate,  and  consequently  the 
northern  limit  of  Siberian  pine  is  considerably  displaced 
southward. 

Siberian  pine  almost  absolutely  escapes  the  moimtain 
massif  of  the  Prepolar  Urals  from  the  west;  its  distribu- 
tion is  restricted  to  the  Pechora  lowland;  in  rare  cases  it 
occurs  in  the  moimtain  part  along  deep  erosion-tectonic 
valleys  of  the  large  rivers. 

On  the  eastern  slope  of  the  Prepolar  Urals,  however, 
Siberian  pine  ascends  rather  high  in  the  mountains,  al- 
most to  the  upper  forest  limit.  In  the  "subgoltsy"  belt 
(analogous  to  the  subalpine  belt),  in  the  region  of  Gorodkov 
Mountain  (Sale-Ur-Oika),  Siberian  pine  shows  dwarfed 
growth  (2.5  to  5  m  in  height),  exhibiting  trunks  thickened 
at  the  basal  portion  of  the  tree  and  tapering  upward,  and 
with  wide-branched  crowns  of  low  stability.  Frequently, 
such  dwarfed  trees  bear  abundant  cones.  At  lower  eleva- 
tions, within  the  mountain-taiga  belt  in  the  plain,  Sibe- 
rian pine  reaches  10  m  in  height.  From  the  eastern  side, 
along  the  foothills,  this  plant  extends  to  the  Polar  Urals 
reaching  66°  10'  N.  latitude  in  the  mountains  of  the  so- 
called  Small  Urals. 

More  intensive  advancement  of  Siberian  pine  north- 
ward, along  the  eastern  slope  of  the  Ural  Mountain  range, 
was  favored  partly  by  historical  causes.  This  species  in- 
vaded from  Siberia  to  new  places  in  the  Urals.  It  arrived 
earher  on  the  eastern  slope  of  the  Ural  Mountain  range, 
favored  mainly  by  the  specific  natural  conditions  of  the 
eastern  slope.  The  eastern  slope  of  the  Prepolar  and 


57 


Polar  Urals  has  a  more  continental  climate  than  the  west- 
ern slope.  Relative  air  humidity  is  lower  here;  annual 
precipitation  including  that  in  the  winter  is  less  heavy; 
and  snow  cover  melts  earlier  in  spring.  Consequently,  the 
growing  season  is  longer,  and  the  top  soil  layer  is  better 
heated. 

On  its  northern  boundaries,  Siberian  pine  is  distributed 
disjimctively  and  occurs  as  dispersed  small  clumps  or  soli- 
tary specimens  in  the  taiga.  All  these  extreme  localities 
are  rather  disjimct  from  each  other.  This  may  be  ex- 
plained primarily  by  the  fact  that  only  in  rare  locations 
are  the  combinations  of  various  environmental  factors 
favorable  for  growth  of  Siberian  pine  at  the  limits  of  its 
range.  On  the  other  hand,  the  dispersion  of  this  species 
has  been  reduced  here  and  there  due  to  forest  fires,  exten- 
sion of  bogs,  trampling  of  seedlings  and  young  growth 
by  reindeer,  and  extermination  by  humans.  In  locations 
where  human  activity  is  reduced  to  a  minimum,  Siberian 
pine  invades  new  places  north  of  its  present  limit.  This 
advance  is  observed  especially  clearly  in  the  mountain 
regions  of  the  Urals. 


THE  SOUTHERN  LIMIT 

The  southern  borderline  of  continuous  distribution  of 
Siberian  pine  (fig.  2)  passes  through  the  Russian  plain 
near  the  city  of  Perm;  then  it  turns  southward  along  the 
western  slope  of  the  middle  Urals,  reaching  the  most 
southern  position  at  56°  10'  N.  latitude;  then  it  switches 
back  north  following  the  eastern  slope  of  the  middle  Urals 
until  retiring  into  western  Siberia. 

A  few  outliers  occur  beyond  this  limit.  The  most  inter- 
esting of  them  are  to  be  found  in  the  mountain  dark  conif- 
erous taiga  of  the  southern  Urals  in  the  region  of  the  town 
of  Zlatoust  and  in  the  Ilmen  Mountains. 

At  the  southern  limit  of  its  range,  Siberian  pine  usually 
grows  in  the  form  of  single-standing,  tall  trees.  Only  on 
rocky  outcrops  does  it  exhibit  prostrated  growth  forms 
(krummholz). 

Natiiral  regeneration  of  Siberian  pine  near  the  southern 
boimdary  of  its  distribution  is  generally  satisfactory.  How- 
ever, there  are  many  factors  unfavorable  to  regeneration, 
such  as  seed  collection  difficulties,  fires,  and  trsimpling  of 


Figure  2— Southern  distribution  limit  of  Siberian  stone  pine  in  the  Urals 
and  in  the  adjacent  plains.  *  =  isolated  occurrences. 


58 


seedlings  by  livestock.  In  locations  where  such  vmfavor- 
able  influences  are  absent,  or  they  manifest  themselves 
to  a  lesser  degree,  regeneration  of  Siberian  pine  is  usually 
good.  So,  for  example,  Siberian  pine  reproduces  very  suc- 
cessfully on  raised  sites  protected  from  fire  within  the 
Bakhmet  Bog,  which  is  surroimded  by  Scots  pine  forests 
in  the  basin  of  the  Pyshma  River. 

Along  the  western  slope  of  the  Urals,  as  well  as  in  cer- 
tain sites  within  the  watershed  zone  where  the  climate  is 
more  humid,  Siberian  pine  has  invaded  new  places  signifi- 
cantly farther  south  than  in  the  Preural  and  Transural  re- 
gions. Its  southern  limit  projects  southward  in  the  form 
of  tongues  that  comprise  the  western  slope  and  part  of  the 
watersheds  of  the  moimtain  range.  Climate  of  the  eastern 
slope  of  the  middle  Urals  is  drier;  the  amoiint  of  precipita- 
tion is  less  compared  to  the  western  slope.  Therefore,  on 
the  eastern  slope  of  the  movmtain  range  and  in  the  foot- 
hills zone  adjacent  to  it  (the  region  of  predominance  of 
Scots  pine  forests),  Siberian  pine  occurs  considerably  far- 
ther north. 

Within  the  peatland  between  the  rivers  of  the  western 
Siberian  plain,  the  southern  limit  of  Siberian  pine  turns 
south  again.  There  Siberian  pine  grows  mainly  on  raised 
sites  within  bogs,  where  it  is  well  protected  from  fire. 

Siberian  pine  invaded  the  southern  Urals  later  than 
spruce  and  fir  {Abies  sibirica)  and  has  not  had  enough 
time  in  the  prehistoric  period  to  reach  the  climatically 
influenced  limit  of  its  distribution.  Human  activities 
hampered  the  southward  advance  of  Siberian  pine  into 
new  places  along  the  Ural  Mountain  range.  Here  and 
there,  numbers  of  this  species  have  declined.  Several  cen- 
turies ago  Siberian  pine  had  been  distributed  farther 
south  in  the  Urals  compared  to  its  contemporary  limit. 
Since  then  it  has  receded  northward  due  to  forest  fires 
and  himian  impact.  There  is  clear  evidence  of  recent  de- 
cline in  total  distribution.  In  the  southern  part  of  the 
area,  Siberian  pine  occurs  sporadically  in  small  groves, 
groups  of  trees,  or  as  single  trees.  It  grows  in  places  least 
endangered  by  fire,  most  frequently  on  moist  soils  in  the 
headwater  regions  and  valleys  of  brooks  and  rivulets,  on 
the  margins  of  bogs  and  on  raised  sites  within  them,  on 
lake  shores  (somewhere  close  to  water),  and  on  steep  out- 
crops of  acid  rocks. 

Formerly,  the  region  of  continuous  distribution  of  Sibe- 
rian pine  extended  south  at  least  to  the  isolated  localities 
mentioned  earlier,  but  probably  this  tree  species  extended 
even  farther  south.  Literature  data  as  well  as  oral  state- 
ments by  local  residents  provide  evidence  of  it.  Siberian 
pine  disappeared  from  these  places  not  only  due  to  cutting 
of  timber  and  forest  fires,  but  also  because  local  inhabit- 
ants picked  all  the  cones,  often  before  cones  were  com- 
pletely ripe.  They  also  dug  up  seedlings  for  transplanting 
into  personal  gardens. 

DISCUSSION  AND  CONCLUSIONS 

The  rather  complicated  picture  of  distribution  of  Sibe- 
rian pine  in  the  Urals  depends  to  a  large  extent  on  cli- 
matic difl'erences  between  the  relatively  himiid  western 
slope  of  the  range  and  the  more  continental  eastern  slope. 


Humidity  of  the  western  slope  restricted  the  northward 
advance  of  Siberian  pine,  but  favored  its  advance  to  the 
south.  On  the  contrary,  continentality  of  the  eastern 
slope  favored  the  advance  of  this  species  northward,  but 
restricted  its  advance  southward. 

At  both  the  northern  and  southern  Umits  of  its  distribu- 
tion, Siberian  pine  does  not  occur  high  up  in  the  moun- 
tains. Its  northernmost  occurrences  in  high  mountains 
are  in  the  eastern  part  of  the  Prepolar  Urals. 

In  the  northern  Urals,  Siberian  pine  is  widespread  al- 
most everywhere  in  the  high-mountain  belts.  On  certain 
mountains  (Oika-Nyor,  Pas-Nyor,  Chistop,  and  others), 
however,  it  forms  the  upper  forest  limit.  High-mountain 
Siberian  pine-dominated  forests  grow  on  steep  slopes  with 
poorly  developed  soils.  Near  the  upper  forest  limit,  such 
forests  may  be  foimd  mainly  in  less  high  mountains  far 
removed  from  large  mountain  massives.  In  such  isolated 
mountains,  the  upper  forest  limit  is  caused  chiefly  by 
edaphic  factors,  while  the  cHmate  of  their  treeless  zone 
and  of  the  subgoltsy  belt  is  less  severe.  Dwarfed,  pros- 
trate, and,  in  most  cases,  noncone-producing  specimens 
of  Siberian  pine  of  the  northern  Urals  occur  in  high- 
moimtain  tundras  significantly  higher  in  elevation  than 
the  normal  upper  forest  limit.  For  instance,  dwarf  speci- 
mens occur  on  the  movmtains  Isherim,  Oika-Chahl, 
Yalping-Nyor,  Denezhkin  Kamen,  Konzhakovsky  Kamen, 
Kosvinsky  Kamen,  and  others.  The  occurrences  of  kmmm- 
holz  growth  forms  above  the  normal  upper  forest  limit 
have  to  be  explained  by  the  dispersal  of  seeds  into  the 
moimtain  tundra  zone  by  the  nutcracker  and  rodents. 

In  the  middle  Urals  near  the  simamits  of  relatively  high 
moimtains,  Siberian  pine  also  forms  singly  standing, 
dwarfed  trees  (the  mountains  Kachkanar,  Starick-Kamen, 
and  others).  South  of  Moimt  Starick-Kamen,  this  species 
is  never  foimd  in  high-movmtain  belts,  but  in  the  forest 
belt  its  area  spreads  considerably  farther  south. 

Judging  by  its  biological  and  ecological  features,  Sibe- 
rian pine  may  not  be  considered  as  a  species  becoming 
extinct.  It  is  perfectly  well  adapted  to  the  taiga  environ- 
ment; it  is  a  good  cone  producer,  and,  if  human  impact 
does  not  interfere,  it  regenerates  quite  satisfactorily.  Al- 
most everywhere,  regeneration  occurs  close  to  old  trees. 
However,  one  should  bear  in  mind  that  Siberian  pine 
grows  more  slowly  than  other  shade-tolerant  coniferous 
trees.  Therefore,  ground  fires  that  annihilate  second- 
growth  trees  adversely  afl'ect  natural  regeneration  of  this 
species. 

The  border  lines  of  Siberian  pine,  especially  on  the 
southern  border,  have  been  reduced  by  careless  and  some- 
times barbarian  treatment  of  this  valuable  plant.  It  is 
necessary  to  secure  proper  protective  measiires  for  its 
southern  localities  in  the  Urals  and  to  use  seeds  gathered 
from  trees  mainly  for  sowing  under  the  forest  canopy  and 
in  tree  nurseries. 

In  the  coTorse  of  intensive  harvest  cuttings  in  the  dark 
coniferous  fir-spruce  forests  that  have  individual  Siberian 
pine  trees  in  the  admixture,  foresters  usually  retain  Sibe- 
rian pine  as  seed  trees  for  natural  regeneration.  How- 
ever, in  many  cases,  this  does  not  work  well  because  in 
the  heavy  clay  and  loamy  soils  of  the  plains,  and  in  the 


59 


mountains  on  poorly  developed  soils,  such  trees  r\in  the 
danger  of  windfall.  To  avoid  windfall,  it  is  necessary  to 
retain  clumps  of  Siberian  pine  with  accompanying  tree 
species  instead  of  single  specimens  of  Siberian  pine. 

Natural  conditions  of  the  middle  Urals  and  of  the 
moiintain-range  portion  of  the  southern  Urals  (at  least 
up  to  54°  N.  latitude)  are  favorable  for  regeneration  and 
growth  of  Siberian  pine.  This  valuable  tree  species  may 
be  reintroduced  to  areas  south  of  its  present  distribution. 
This  can  be  concluded  from  the  successful  cultivation  of 
Siberian  pine  south  of  its  present  border  line  in  the  moun- 
tain taiga  regions  of  the  middle  and  southern  Urals. 


REFERENCES 

Gorchakovsky,  P.  L.  1975.  The  plant  world  of  the  high- 
mountain  Urals.  Moscow:  "Nauka"  Publishing.  284  p. 
[In  Russian]. 

Gorchakovsky,  P.  L.  1989.  Horizontal  and  altitudinal 
differentiation  of  the  vegetational  cover  of  the  Ural 
Mountains.  Pirineos,  N  133:  33-54. 

Gorodkow,  B.  N.  1929.  The  Polar  Urals  in  the  upper 
reaches  of  the  rivers  Voykar,  Synya  and  Lyapin.  In: 
Materials  of  the  Committee  for  Expeditional  Re- 
searches, V.  7,  Leningrad:  1-28.  [In  Russian]. 

Tyrtikov,  A.  P.  Growth  of  tree  roots  at  the  northern  limit 
of  forests.  Bull,  of  Moscow  Soc.  of  Naturalists.  Section 
of  Biology.  59(1):  72-82.  [In  Russian]. 


60 


ECOPHYSIOLOGICAL  INVESTIGATIONS 
ON  CEMBRAN  PINE  AT  TIMBERLINE  IN 
THE  ALPS,  AN  OVERVIEW 

Rudolf  Hasler 


Abstract — During  the  nineteenth  century  foresters  as  well  as 
scientists  became  interested  in  learning  more  about  cembran 
pine  (Pinus  cembra)  in  the  Alps,  but  extensive  ecophysiological 
fieldwork  did  not  begin  before  the  1930's.  Great  progress  was 
made  after  the  estabUshment  of  permanent  field  stations  at  the 
timberUne  at  Obergurgl  (1953),  Patscherkofel  near  Innsbruck 
(1963),  and  Stillberg  near  Davos  (1959).  In  a  brief  overview, 
some  selected  results  of  ecophysiological  research  in  cembran 
pine  are  presented. 


The  old  and  bizarre  cembran  pines  (Pinus  cembra  L.), 
which  form  the  timbedine  over  a  wide  range  in  the  Cen- 
tral Alps,  have  always  fascinated  himaans.  In  earlier 
days,  the  tree  was  of  some  importance.  It  produced  not 
only  wood  but  also  nuts  that  could  be  used  as  food.  There- 
fore, this  tree  has  been  of  some  interest  for  a  long  time, 
and  a  lot  of  books  and  papers  about  cembran  pine  in  the 
Alps  have  been  published,  in  forestry  as  well  as  in  biologi- 
cal science. 

This  short  overview  presents  only  some  selected  ex- 
amples of  the  ecophysiological  research,  with  a  special 
focus  on  gas  exchange.  A  general  overview  on  ecophysio- 
logical problems  at  the  timberline  was  published  by 
TranquilUni  (1979).  Many  other  subjects,  such  as  growth, 
genetics,  and  mycorrhiza,  have  been  investigated  in 
Austria,  France,  Italy,  and  Switzerland  but  are  not  men- 
tioned in  this  overview. 

BEGINNING  OF  INVESTIGATIONS 

In  the  middle  of  the  nineteenth  century,  foresters  be- 
came interested  in  the  cembran  pines  growing  in  the  Alps. 
Fankhauser  (1853)  described  the  distribution  of  these 
trees  together  with  the  climatological  conditions  they 
need  for  growing  well.  He  mentioned  that  it  is  not 
possible  to  transfer  this  tree  species  to  low  elevations. 
Kemer  (1866)  and  Simony  (1870)  pubhshed  similar  data 
for  parts  of  Austria.  At  the  same  time,  the  diseases  and 
the  insect  enemies  of  the  cembran  pine  in  the  Alps  at- 
tained some  importance.  They  have  been  described  by 
Keller  (1890)  and  others. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Rudolf  Hasler  is  Head  of  the  Section  "Forest  and  Climate,"  Swiss  Fed- 
eral Institute  for  Forest,  Snow  and  Landscape  Research,  CH-8903 
Birmensdorf,  Switzerland. 


ECOLOGICAL  AND  GEOGRAPfflCAL 
OBSERVATIONS 

More  and  more  scientists  became  interested  in  cembran 
pine  at  the  end  of  the  nineteenth  century  and  the  early 
twentieth  century  to  learn  more  about  these  trees,  which 
grow  on  the  most  extreme  sites  in  the  mountains.  Mono- 
graphs concerning  cembran  pine  in  Switzerland  (RikH 
1909)  and  in  Austria  (Nevole  1914)  were  pubhshed.  Many 
other  publications  were  written  at  this  time.  The  main 
subjects  were  growth,  and  again  fungal  diseases  and  prob- 
lems with  insects  (Fankhauser  1903;  Keller  1901,  1910; 
and  others). 

FIRST  ECOPHYSIOLOGICAL  FIELD 
EXPERIMENTS 

Ecophysiological  investigation  in  the  field  began  in  the 
thirties  on  Patscherkofel  near  Innsbruck  (Austria).  To 
work  near  the  timberline  in  the  Alps  was  a  logistic  prob- 
lem. The  cembran  pines  only  grow  near  the  timberline, 
at  high  altitudes.  It  was  extremely  hard  to  work  there  in 
earher  times. 

Cartellieri  (1935)  was  probably  the  first  to  measiu*e 
photosynthesis  and  water  relations  throughout  a  whole 
season  in  cembran  pine  at  timberline.  At  the  same  time 
he  also  took  measurements  of  different  dwarf  shrubs  at 
high  elevation.  He  measured  photosynthesis  by  absorbing 
the  carbon  dioxide  in  barite-water,  and  then  he  deter- 
mined the  COg  assimilation  by  titration  in  the  laboratory. 
Some  idea  about  the  difficulties  of  working  at  the  timber- 
line  in  those  days  is  given  in  his  papers:  missing  data  in 
the  annual  course  are  explained  as  "walking  back  to  the 
laboratory  I  lost  some  of  the  probes  because  some  of  the 
glass  flasks  were  broken,"  or  in  another  example,  "at  this 
time  I  got  ill  and  had  to  stop  the  investigation."  Never- 
theless his  results  show  more  or  less  the  same  values  as 
we  are  measuring  today  with  ovu*  modem  climatized  and 
very  expensive  instrumentation.  Also  about  the  same 
time  Ulmer  (1937)  investigated  the  annual  course  of  fi*ost 
hardiness  and  measured  in  parallel  the  osmotic  values 
in  the  needles. 

PERMANENT  FIELD  STATIONS  AND 
CONTINUOUS  CO2  MEASUREMENTS 

Great  progress  in  timberline  research  in  the  Alps  was 
made  in  the  1950's.  Again  in  Austria,  a  permanent 
experimental  field  station  at  an  altitude  of  about  1,950  m 
above  sea  level  (a.s.l.)  was  established  in  1953  near 


61 


Obergurgl.  At  the  same  time,  the  newly  developed  infra- 
red gas  analyzer  (IRGA)  made  it  possible  to  measm'e 
continuously  carbon  dioxide  in  the  air. 

Tranquillini  (1955)  conducted  experiments  at  the  new 
field  station  at  Obergurgl  to  investigate  the  relations  be- 
tween light,  temperature,  and  photosynthesis  in  needles 
of  young  cembran  pines  growing  in  the  sun  or  in  the  shad- 
ows. He  determined  the  light  compensation  points  for 
these  needles  as  500  and  150  Lux,  respectively,  and  the 
light  saturation  points  as  25  and  18  kLux,  respectively. 
The  COg  assimilation  values  he  reported  reached  2.5  mg 
and  2.1  mg  COJg  dry  weight,  respectively.  Net  photo- 
synthesis starts  at  -5  °C  and  has  an  optimum  range  be- 
tween 10  and  15  °C.  The  upper  temperature  limit  is  at 
35  °C. 

The  microclimatic  conditions  during  wintertime  and 
needle  and  soil  temperatures  as  well  as  gas  exchange  and 
water  relations  in  young  cembran  pines  growing  on  natu- 
ral sites  at  the  timberline  were  published  by  Tranquillini 
(1957).  He  showed  that  in  wintertime,  even  in  warmer 
weather  conditions,  no  net  photosynthesis  was  detectable. 
Holzer  (1958)  explained  this  lack  of  photosynthesis  by  a 
changed  cell  structure  as  well  as  by  a  changed  physiolo- 
gical behavior  during  wintertime. 

Pisek  and  Winkler  (1958)  measured  the  annual  course 
of  the  capacity  of  net  photosynthesis  and  of  respiration  in 
detached  shoots  of  cembran  pines  growing  at  different  al- 
titudes. During  wintertime,  they  found  an  always  nega- 
tive carbon  dioxide  balance  in  the  trees  at  the  timberline, 
whereas  those  growing  at  lower  elevations  could  produce 
a  positive  one,  at  least  during  periods  with  temperatures 
above  the  freezing  point.  Keller  (1970)  confirmed  these 
findings,  using  another  method.  Tranquillini  (1958, 
1963d)  investigated  the  annual  course  of  frost  resistance. 
He  found  that  young  trees  are  more  sensitive  to  frost  than 
mature  trees.  Young  trees  become  damaged  by  tempera- 
tures between  -15  to  -30  °C,  while  mature  trees  can  sur- 
vive temperatures  lower  than  —45  °C.  Keller  and  Beda- 
Puta  (1973)  analyzed  stem  respiration  in  cembran  pine 
during  wintertime. 

Pisek  and  Winkler  (1959)  studied  photosynthesis  in  re- 
lation to  temperature  in  different  light  conditions  as  well 
as  at  different  altitudes.  With  increasing  light  intensity, 
the  temperature  optimum  is  shifted  to  higher  tempera- 
tures. There  are  some  difficulties  in  comparing  all  these 
older  results  with  the  measurements  of  today,  because  not 
all  parameters  can  be  transformed  (for  example,  illumina- 
tion [Lux]).  Some  other  values,  such  as  the  needle  sur- 
faces, were  normally  not  used  as  a  reference  for  gas  ex- 
change measurements  before  the  1980's. 

A  COg  budget  for  a  whole  year  was  estimated  by 
TranquiUini  (1959a,  1959b,  1963a,  1963b).  Total  CO^ 
uptake  per  year  (gross  photosynthesis)  of  a  5-  to  8-year- 
old  tree  at  2,000  m  a.s.l.  was  about  7.8  kg.  About  38  per- 
cent was  lost  by  respiration  from  needles,  branches,  stem, 
and  roots.  According  to  the  measured  net  COg  assimi- 
lation, a  tree  should  produce  2.2  g  dry  matter  per  gram  of 
needles.  In  reality  they  produced  0.65  g.  The  author 
mentioned  that  most  of  the  missing  carbon  dioxide  may 
have  been  used  by  the  mycorrhiza. 

In  the  second  half  of  the  1950's,  ecological  investiga- 
tions with  the  aim  of  finding  methods  for  successful 


afforestation  at  high  altitude  were  started  in  Switzerland 
(Stillberg,  Davos:  2,000-2,230  m  a.s.l.)  and  in  Austria 
(Haggen,  Sellraintal:  1,730-1,900  m  a.s.l.).  Cembran  pine 
was  one  of  the  species  used  in  these  field  experiments, 
which  still  are  going  on.  It  takes  a  long  time  to  get  results 
from  trees  growing  at  high  elevation  near  the  timberline. 
Svmimaries  containing  prelimin£iry  results  were  published 
by  Schonenberger  and  Frey  (1988)  for  Stillberg  and  by 
Kronfuss  (1980,  1983,  1986)  for  Haggen. 

LABORATORY  CLIMATE  CONTROL 

The  technical  possibilities  developed  fast.  On 
Patscherkofel  (1,950  m  a.s.l.)  a  climate  laboratory  was 
established  in  1963.  In  this  laboratory  a  device  was  in- 
stalled that  allows  analysis  of  the  gas  exchange  of  yoimg 
trees  in  climatized  windchannels  at  high  altitude.  Now  it 
became  possible  to  control  exactly  the  climatic  parameters 
such  as  temperature  and  air  and  soil  humidity  as  well  as 
windspeed,  which  was  a  brand-new  innovation.  Using 
this  new  instrumentation  old  measurements  could  be  ve- 
rified and  a  lot  of  new  possibilities  were  opened  up.  In 
particular,  water  relations  could  now  be  investigated  un- 
der controlled  climatic  conditions. 

Tranquillini  (1963c)  investigated  the  CO^  assimilation 
in  relation  to  air  and  soil  humidity.  In  water-saturated 
soil,  dry  air  (25  percent  relative  hmnidity)  reduced  photo- 
synthesis in  cembran  pines  by  about  33  percent.  A  little 
less  water  in  the  soil  (half  of  field  capacity)  allowed  some 
higher  photosjnithetic  rates.  Resistance  to  dry  himiidity 
increases  from  spring  to  autimin  (Tranquillini  1965). 

Pisek  and  others  (1967,  1968, 1969)  explored  the  tem- 
perature limits  of  photosynthesis  to  chill  and  frost,  and  to 
high  temperatures.  COg  assimilation  is  operative  between 
-4.7  and  +36  °C,  but  the  trees  can  survive  temperatures 
between  below  -50  and  up  to  48  °C.  Of  great  importance 
is  not  only  the  time  factor,  but  also  the  physiological  con- 
dition of  the  plant.  Temperature  hardiness  changes  with 
the  seasons,  £md  young  trees  or  seedlings  are  more  sensi- 
tive to  extreme  temperatures  than  mature  trees.  The  re- 
lation between  photosynthetical  capacity  and  temperature 
was  analyzed  by  Tranquillini  and  Machl-Ebner  (1971). 
Another  factor  investigated  was  day  length.  Bamberg 
and  others  (1967)  and  Schwarz  (1970a,  1970b,  1971)  have 
shown  that  day  length  does  not  influence  the  photosyn- 
thetic  capacity  during  winter  depression,  but  trees  kept 
in  warm  conditions  during  winter  commenced  growth  in 
spring  much  earlier  than  plants  in  nature.  The  variation 
in  frost  resistance  of  the  trees  changed  in  parallel  with 
the  temperature.  In  long-day  regimes,  frost  resistance  of 
cembran  pine  remained  at  a  lower  level  than  those  grow- 
ing imder  short-day  conditions.  The  researchers  supposed 
that  an  endogenous  rhythm  also  influences  resistance  to 
extreme  temperatures. 

Caldwell  (1970),  using  the  climatized  wind  tunnel  of  the 
Patscherkofel  laboratory,  has  shown  the  effect  of  wind  on 
stomatal  aperture,  photosynthesis,  and  transpiration. 
He  did  not  detect  a  great  influence  of  low  windspeeds  on 
these  parameters.  At  higher  windspeeds  (up  to  8  m/sec) 
photosynthesis  and  transpiration  were  slightly  reduced. 
Baig  and  others  (1974)  and  Baig  and  Tranquillini  (1980) 
investigated  the  cuticular  transpiration  rate  of  cembran 


62 


pine  and  found  that  cuticular  transpiration  in  this  tree 
is  very  slow  but,  as  in  other  species,  cuticular  resistance 
decreases  in  paredlel  to  the  elevation  at  which  the  shoots 
were  collected. 

Havranek  (1972)  found  that  cold  soil  temperatures 
reduce  photosynthesis  and  transpiration  in  cembran  pine. 
Havranek  and  Benecke  (1978)  investigated  the  influence 
of  soil  moistiire  on  transpiration  and  photosynthesis. 
Cembran  pine  begins  to  close  the  stomata  at  a  soil  water 
potential  of  about  -0.4  bars.  Compared  to  other  species, 
cembran  pine  maintains  a  high  water-use  efficiency  and 
uses  limited  soil  water  economically.  Giinthardt  and 
Wanner  (1982)  showed  that  needles  from  timberline  trees 
have  on  the  average  more  wax  than  those  grown  at  a 
lower  elevation. 


Another  step  to  get  more  precise  resvilts  for  the  CO^ 
balance  of  cembran  pine  was  taken  by  Tranquillini  and 
Schiitz  (1970).  They  measured  bark  respiration  at  the 
timberline  and  reported  that  b£irk  respiration  also  shows 
an  annual  course.  In  general,  bark  respiration  of  cem- 
bran pine  was  rather  high  compared  to  that  of  other  spe- 
cies. The  loss  of  carbon  dioxide  by  stem  respiration 
throughout  a  year  was  calculated  to  be  about  18.5  percent 
of  gross  photosynthesis.  Measuring  stem  respiration  con- 
tinuously, Havranek  (1981)  found  that  there  is  a  marked 
fluctuation  in  stem  respiration  during  the  growing  season, 
but  he  could  not  find  any  relation  between  activity  of  stem 
respiration  and  tree  ring  growth. 


Net  photosynthesis 

40- 

■a  32" 
E  2411 

o  16 
E  81 
E 

-8 


I  I  I  I   I  B  I  I  I   I   I  11  I  I  I   I  I  11  I  I  I  I  I  11  I  I  I  I  I   11  I  I   I   I  I  I  I   I   I  I  I  11  I   I   I  I  I  H  I  I  I  I  I  11  I  I  I 

5  15  2515  15  2515  15  2515  15  25  15  15  25  1 5  15  251  5  15  2515  15  2515  15  25  I  5  15 
Photosynthetically  active  radiation 


'l  I  I  I  I  I  11  I  I  I  I  I  11  rr  I  I  I  11  I  I'V  ri  iTTi  i  i  i  ii  T  i  i  i  i  i  i  i  i  i  i-ir  i  i  i  i  i  n  i  i  i  i  i  ii  i  i  r 
15  15  2515  15  2515  15  251  5  15  25  I  5  15  25  I  5  15  2515  15  25  I  5  15  2515  15  25  I  5  15 

Air  temperature 


5  15  2515  15  2515  15  251  5  15  25  I  5  15  25  1 5  15  2515  15  2515  15  251  5  15  25  I  5  15 


Soil  temperature 


I  M  III  I  I  I  I  II  I  I  I  I  I  II  I  I  I  I  I  11  I  I  I  I  I  11  I  I  I  I  I  I  I  (  I  I  I  11  I  I  I  I  I  11  I  I  I  I  I  11  I  I  !' 

1 5  15  2515  15  25  1 5  15  2515  15  25  1 5  15  2515  15  251  5  15  25  I  5  15  2515  15  2515  15 


Water  vapor  pressure  difference 


Is  1525'! 


15  15  2515  15STr'^l525 


Sep  Oct 


Nov 


5  15  25  15  15  2515  15  251 

Dec     Jan      Feb  Mar 


5  15 

Apr     May  Jun 


25 
20 
15 
10 
5 
0 
-5 
-10 
-15 
-20 


Figure  1 — Photosynthesis  in  cembran  pine, 
photosynthetically  active  radiation,  air  and  soil 
termperature,  and  air  humidity  measured  at  the 
timberline  (Stillberg,  1,980  m  a.s.l.,  1979-80). 


63 


FIELD  CLIMATE  CONTROL 

In  the  late  sixties,  the  thermoelectrically  climatized  gas 
exchange  chambers  were  developed  by  Koch  and  others 
(1968).  Hasler  and  Blaser  (1981)  measured  the  first  daily 
courses  in  young  cembran  pines  on  the  east-  and  the 
north-facing  slopes  of  an  avalanche  gully  at  the  timberline 
at  Stillberg.  The  two  sites  exemplify  marked  differences 
in  microclimate  (mean  values  of  soil  surface  temperatures 
per  growing  season  differ  by  more  than  10  °C).  These 
differences  in  microclimate  also  £iffect  the  photosynthetic 
rates  of  afforested  cembran  pines.  Carbon  dioxide  gain, 
summarized  over  a  season,  of  the  trees  on  the  north- 
facing  slope  was  34  percent  less  them  for  those  on  the 
east-facing  slope  (Turner  and  Streule  1983). 

Gas  exchange  measured  in  combination  with  meteoro- 
logical parameters  was  registered  in  a  branch  of  an  older 
cembran  pine  at  Stillberg  (1,980  m  a.s.l.)  throughout 
a  whole  winter  season  in  1979-80  (fig.  1).  Havranek 
(1981)  published  an  annual  course  of  growth  and  a 
carbon  dioxide  balance,  taken  near  the  Patscherkofel 
laboratory.  During  summer  the  measurements  were 
taken  with  climatized  chambers. 

In  the  middle  of  the  1980's  the  so  called  "forest  decline" 
had  high  priority  in  forest  research  in  Europe.  The  main 
work  in  this  field  was  done  on  other  tree  species  than 
cembran  pine.  Nevertheless,  some  investigations  were 
made  on  this  species.  Havranek  (1987)  took  measure- 
ments on  cembran  pine  at  the  timberline  to  investigate 
stress  caused  by  air  pollution  in  combination  with  climatic 
factors.  Genys  and  Heggestad  (1978)  did  not  find  a  high 
susceptibility  of  cembran  pines  to  ozone  and  sulphur  di- 
oxide, but  MinarCic  and  KubiCek  (1991)  suggested  that 
dust  could  influence  the  stomata  in  needles  of  cembran 
pines  and  thus  affect  photosynthesis.  Lutz  and  others 
(1988)  emalyzed  the  photos)nithetic  pigments  of  cembran 
pines  growing  at  different  altitudes,  which  thus  are  ex- 
posed to  different  ozone  concentrations.  Nebel  and  Matile 
(1992)  investigated  the  cause  of  uncommon  yellowing  of 
the  needles  of  cembran  pines  by  analyzing  photosynthe- 
sis, nutrient  relations,  longevity,  and  senescence  of  these 
declining  needles. 

The  newest  measm*ements  in  cembran  pine  from  the 
Stillberg  experimental  station  are  presented  in  these  pro- 
ceedings (Koike  and  others).  These  are  probably  the  first 
measvirements  of  photosynthesis  in  cembran  pine  at  the 
timberline  with  an  eirtificially  elevated  atmospheric 
concentration.  A  detailed  publication  giving  these  results 
is  in  preparation  (Koike  and  others  1993). 

CONCLUSIONS 

At  the  beginning  of  the  research  on  cembran  pine,  about 
150  years  ago,  scientists  worked  mainly  at  the  forest 
level.  With  the  development  of  sophisticated  measure- 
ment instruments,  branches,  twigs,  or  even  needles  were 
the  focus  of  interest.  Rarely  did  anyone  try  to  extrapolate 
these  results  from  the  small  parts  to  a  whole  tree,  a  prob- 
lem which  has  still  not  been  satisfactorily  resolved.  Prob- 
ably we  should  try  to  go  back  directly  to  the  level  of  whole 
trees  or  even  of  stands,  especially  when  we  want  to  under- 
stand changes  caused  by  external  influences,  such  as  air 


pollution,  rising  carbon  dioxide,  changing  UV  radiation, 
and  many  others. 

ACKNOWLEDGMENT 

I  thank  Mrs.  M.  J.  Sieber  for  critical  editing  of  the 
English  version  of  my  manuscript. 

REFERENCES 

Baig,  M.  N.;  Tranquillini,  W.  1980.  The  effect  of  wind  and 
temperature  on  cuticular  transpiration  of  Picea  abies 
and  Pinus  cembra  and  their  significance  in  desiccation 
damage  at  alpine  treeline.  Oecologia.  47:  252-256. 

Baig,  M.  N.;  Tranquillini,  W.;  Havranek,  W.  M.  1974. 
Cuticulare  Transpiration  von  Picea  abies-  und  Pinus 
cembra-Zvfeigen  aus  verschiedener  Seehohe  imd  ihre 
Bedeutung  fiir  die  winterliche  Austrocknung  der 
Baiune  an  der  alpinen  Waldgrenze.  Centredblatt  fiir 
das  gesamte  Forstwesen.  91:  195-211. 

Bamberg,  S.;  Schwarz,  W.;  TranquiUini,  W.  1967.  Influen- 
ce of  daylength  on  photosynthetic  capacity  of  stone  pine 
(Pinus  cembra  L.).  Ecology.  48:  264-269. 

Caldwell,  M.  M.  1970.  The  effect  of  wind  on  stomatal 
aperture,  photosynthesis,  and  transpiration  of  Rhodo- 
dendron ferrugineum  L.  and  Pinus  cembran  L. 
Centralblatt  fur  das  gesamte  Forstwesen.  87:  193-201. 

Cartellieri,  E.  1935.  Jahresgang  von  osmotischem  Wert, 
Transpiration  und  Assimilation  einiger  Ericaceen  der 
alpinen  Zwergstrauchheide  und  von  Pinus  cembran. 
Jahrbiicher  fiir  wissenschaftliche  Boteuiik.  82:  460-506. 

Fankhauser,  F.  1853.  Uber  das  Vorkommen  und 
Gedeihen  der  Arve.  Schweizerisches  Forst  Journal. 
S83. 

Fankhauser,  F.  1903.  Der  Kiefemschiittepilz  an  der  Arve. 
Schweizerische  Zeitschrifl  fiir  Forstwesen.  54:  321-323. 

Genys,  J.  B.;  Heggestad,  H.  E.  1978.  Susceptibility  of  dif- 
ferent species,  clones  and  strains  of  pines  to  acute  in- 
jury caused  by  ozone  and  sulfur  dioxide.  Plant  Disease 
Reporter.  62(8):  687-691. 

Gtinthardt,  M.  S.  ;  Wanner,  H.  1982.  Die  Menge  des 
cuticularen  Wachses  auf  Nadeln  von  Pinus  cembran 
(L.)  und  Picea  abies  (L.)  Karsten  in  Abhangigkeit  von 
Nadelalter  und  Standort.  Flora.  172:  125-137. 

Hasler,  R.;  Blaser,  P.  1981.  Nettophotosjmtheseaktivitat 
von  Arve  imd  Larche  auf  verschiedenen  Standorten  der 
subalpinen  Stufe.  Mitteilungen  der  forstlichen 
Bundesversuchsanstalt,  Wien.  140:  179-188. 

Havranek,  W.  1972.  The  influence  of  soil  temperature  on 
photosynthesis  and  transpiration  of  young  trees  and  on 
dry-matter  production,  at  the  timber  line.  Angewandte 
Botanik.  46:  101-116. 

Havranek,  W.  M.  1981.  Stammatmung,  Dickenwachstum 
und  Photosynthese  einer  Zirbe  {Pinus  cembra)  an  der 
Waldgrenze.  Mitteilimgen  der  forstlichen  Bundes- 
versuchsanstalt, Wien.  142:  443-467. 

Havranek,  W.  M.  1987.  Physiologische  Reaktionen  auf 
Klimastress  bei  Baumen  an  der  Waldgrenze.  Sympo- 
smm  Klima  und  Witterung  in  ZusEunmenhang  mit  den 
neuartigen  Waldschaden.  Tagungsbeitrage,  Miinchen: 
115-138. 


64 


Havranek,  W.  M.;  Benecke,  U.  1978.  The  influence  of  soil 
moistiire  on  water  potential,  transpiration  and  photo- 
synthesis of  conifer  seedlings.  Plant  and  Soil.  49(1): 
91-103. 

Holzer,  K.  1958.  Die  winterlichen  Veranderungen  der 
Assimilationszellen  von  Zirbe  (Pinus  cembra  L.  )  und 
Fichte  (Picea  excelsa  Link.  )  an  der  alpinen  Waldgrenze. 
osterreichische  botanische  Zeitschrift.  105:  323-346. 

Keller,  C.  1890.  Tierische  Forstbeschadigiingen  an  der 
Arve.  Osterreichische  Forstzeitung:  267-268. 

Keller,  C.  1901.  Die  Arven-Erkrankungen  im 
Oberengadin.  Schweizerische  Zeitschrift  fur  Forst- 
wesen.  52:  293-299. 

Keller,  C.  1910.  Die  tierischen  Feinde  der  Arve. 
Schweizerische  Centralanstalt  fur  forstliches.  Versuchs- 
wesen,  Mitteilungen.  10(1):  1-50. 

Keller,  T.  1970.  Uber  die  Assimilation  einer  jungen  Arve 
in  Winterhalbjahr.  Biindnerwald.  23(2):  49-54. 

Keller,  T.;  Beda-Puta,  H.  1973.  Gas  exchange  of  leafless 
shoots  in  winter.  Schweizerische  Zeitschrift;  fiir 
Forstwesen.  124(6):  433-441. 

Kemer,  A.  1866.  Wachstumsbedingungen  der  Zirbe. 
osterr.  Monatsschr.  Forstwesen.  1-15. 

Koch,  W.;  Klein,  E.;  Walz,  H.  1968.  Neuartige 
Gaswechsel-Messanlage  fiir  Pflanzen  in  Laboratorium 
und  Freiland.  Siemens  Zeitschrift;.  42:  392-404. 

Koike,  T.;  Hasler,  R. ;  Matyssek,  R.;  Item,  H.  1993.  Sea- 
sonal changes  in  the  photosynthetic  capacity  of  Larix 
decidua  and  Pinus  cembran  planted  on  contrasting 
slopes  at  the  timberline  at  Stillberg  (Davos),  eastern 
Switzerland.  [In  preparation]. 

Kronfiiss,  H.  1980.  Das  Bestandesklima  einer 
HochlagenaufForstung  auf  einem  Sonnenhang.  In: 
Beitrage  zur  subalpinen  Waldforschung.  2.  Folge. 
Mitteilvmgen  der  forstlichen  Bundesversuchsanstalt, 
Wien.  129:  81-103. 

Kronfuss,  H.  1983.  Das  Hohenwachstum  einer  Zirben- 
aufforstung  auf  einem  Siidhang  in  klimatologischer 
Sicht.  Allgemeine  Forstzeitung.  94(12):  330-332. 

Kronfuss,  H.  1986.  Die  Abhangigkeit  des  Hohen- 
zuwachses  der  Zirbe  von  der  Seehohe.  Jahresbericht 
1985.  Forstliche  Bundesversuchsanstalt,  Wien:  137-144. 

Lutz,  C;  Schulte-Hostede,  S.;  Kirchner,  M.;  Reuther  M. 
1988.  Photosjnithetische  Pigmente  aus  Nadelbaumen 
unterschiedlicher  Hohenstufen  des  Otztals.  Internatio- 
nales Symposium  ''Verteilung  und  Wirkimg  von  Photo- 
oxidantien  im  Alpenraum.  Sitzimgsberichte  Miinchen 
1988:  415-425. 

MinarCic,  P.;  KubfCek,  F.  1991.  Localization  of  emissions 
on  the  needle  surface.  1.  Methodology,  morphological 
characterization  of  the  stomata  {Picea  abies,  Pinus 
cembran).  Ekologia,  CSFR.  10(4):  405-413. 

Nebel,  B.;  Matile,  Ph.  1992.  Longevity  and  senescence 
of  needles  in  Pinus  cembran.  Trees.  6(3):  156-161. 

Nevole,  J.  1914.  Die  Verbreitimg  der  Zirbe  in  der  osterr.  - 
ungar.  Monarchic.  Verlag  W.  Frick  Wien.  89  p. 

Pisek,  A.;  Winkler,  E.  1958.  Assimilationsvermogen  und 
Respiration  der  Fichte  (Picea  excelsa  Lind. )  in  verschie- 
dener  Hohenlage  und  der  Zirbe  (Pinus  cembran  L.  )  an 
der  alpinen  Waldgrenze.  Planta.  51:  518-543. 

Pisek,  A.;  Winkler,  E.  1959.  Licht-  und  Temperaturabhan- 


gigkeit  der  COg-Assimilation  von  Fichte  (Picea  excelsa 
Lind. ),  Zirbe  (Pinus  cembran  L. )  und  Sonnenblume 
(Helianthus  annuus  L.  ).  Planta.  53:  532-550. 

Pisek,  A.;  Larcher,  W.;  Moser,  W.;  Pack,  I.  1969. 
Kardinale  Temperaturbereiche  der  Photosynthese 
und  Grenztemperaturen  des  Lebens  der  Blatter  ver- 
schiedener  Spermatophyten.  III.  Temperatur- 
abhangigkeit  und  optimaler  Temperaturbereich  der 
Netto-Photosynthese.  Flora  B.  158(3):  608-630. 

Pisek,  A.;  Larcher,  W.;  Unterholzner,  R.  1967.  Kardinale 
Temperaturbereiche  der  Photosynthese  und  Grenz- 
temperaturen des  Lebens  der  Blatter  verschiedener 
Spermatophyten.  I.  Temperaturminimum  der  Netto- 
assimilation,  Gefrier-  und  Frostschadenbereiche  der 
Blatter.  Flora  B.  157(1):  239-264. 

Pisek,  A.;  Larcher,  W.;  Unterholzner,  R.;  Pack,  I.  1968. 
Kardinale  Temperaturbereiche  der  Photosynthese  und 
Grenztemperaturen  des  Lebens  der  Blatter  verschie- 
dener Spermatophyten.  II.  Temperaturmaximima  der 
Netto-Photosynthese  und  Hitzeresistenz  der  Blatter. 
Flora  B.  158(2):  110-128. 

Rikli,  M.  1909.  Die  Arve  in  der  Schweiz.  Neue  Denkschr. 
Schweizerische  naturforschende  Gesellschaft.  44.  455  p. 

Schonenberger,  W.;  Frey,  W.,  eds.  1988.  Untersuchungen 
zur  Okologie  imd  Technik  der  Hochlagenauffbrstimg. 
Forschimgsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fiir  Forstwesen. 
139(9):  735-820. 

Schwarz,  W.  1970a.  Der  Einfluss  der  Photoperiode  auf 
das  Austreiben,  die  Frostharte  und  die  Hitzeresistenz 
von  Zirben  und  Alpenrosen.  Flora.  159:  258-285. 

Schwarz,  W.  R.  1970b.  Frost  hardiness,  resistance  to  heat, 
and  COg  uptake  in  seedlings  and  mature  stages  of 
Pinus  cembran  L.  Proc.  3.  Forest  Microclimate  Sympo- 
sium, Kananaskis  Forest  Experimental  Station;  1969; 
Seebe,  AB:  137-145. 

Schwarz,  W.  1971.  The  photos)nithetic  capacity  of  some 
evergreens  during  the  winter  and  the  speed  of  their 
recovery  of  activity  after  sharp  frosts.  Berichte  der 
deutschen  botanischen  Gesellschaft.  84(10):  585-594. 

Simony,  F.  1870.  Die  Zirbe.  Jahrbuch  des  osterreichischen 
Alpenvereins.  6:  349-359. 

Tranquillini,  W.  1955.  Die  Bedeutimg  des  Lichtes  und  der 
Temperatur  fiir  die  Kohlensaureassimilation  von  Pinus 
cembran-tJungwuchs  an  einem  hochalpinen  Standort. 
Planta.  46:  154-178. 

Tranquillini,  W.  1957.  Standortsklima,  Wasserbilanz  und 
COg-  Gaswechsel  junger  Zirben  (Pinus  cembran)  an  der 
alpinen  Waldgrenze.  Planta.  49:  612-661. 

Tranquillini,  W.  1958.  Die  Frostharte  der  Zirbe  imter 
besonderer  Beriicksichtigung  autochoner  imd  aus 
Forstgarten  stammender  Jimgpflanzen. 
Forstwissenschaftliches  Centralblatt.  77:  89-105. 

Tranquillini,  W.  1959a.  Die  Stoffproduktion  der  Zirbe  an 
der  Waldgrenze  wahrend  eines  Jahres.  Planta.  54: 
107-129. 

TranquiUini,  W.  1959b.  Die  Stoffproduktion  der  Zirbe 
(Pinus  cembran  L. )  an  der  Waldgrenze  wahrend  eines 
Jahres.  II.  Zuwachs  imd  COg-Bilanz.  Planta.  54(2): 
130-151. 


65 


Tranqmllini,  W.  1963a.  Der  Jahresgang  der  CO^- 
Assimilation  junger  Zirben.  Mitteilungen  der  forst- 
lichen  Bundesversuchsanstalt,  Wien.  60:  501-534. 

Tranquillini,  W.  1963b.  Die  COg-Jahresbilanz  und  die 
Stoifproduktion  der  Zirbe.  Mitteilungen  der  forstlichen 
Biindesversuchsanstalt,  Wien.  60:  535-546. 

Tranquillini,  W.  1963c.  Die  Abhangigkeit  der  Kohlen- 
saureassimilation  junger  Larchen,  Fichten  und  Zirben 
von  der  Luft-  und  Bodenfeuchte.  Planta.  60:  70-94. 

Tranquillini,  W.  1963d.  Uber  die  Frostresistenz  der  Zirbe. 
Mitteilungen  der  forstlichen  Bundesversuchsanstalt, 
Wien.  60:  547-562. 

Tranquillini,  W.  1965.  Uber  den  Zusammenhang  zwischen 
Entwicklungszustand  und  Diirreresistenz  junger  Zirben 
{Pinus  cembran  L.)  im  Pflanzgarten.  Mitteilungen  der 
forstlichen  Bundesversuchsanstalt,  Wien.  66:  241-271. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Berlin:  Springer- Verlag.  137  p. 


Tranquillini,  W.;  Machl-Ebner,  I.  1971.  Uber  den  Einfluss 
von  Warme  auf  das  Photosynthesevermogen  der  Zirbe 
(Pinus  cembran  L.)  und  der  Alpenrose  (Rhododendron 
ferrugineum  L.)  im  Winter.  Report  of  the  Kevo  Subarc- 
tic Research  Station.  8:  158-166. 

TranquilUni,  W.;  Schiitz,  W.  1970.  Uber  die  Rinden- 
atmimg  einiger  Baume  an  der  Waldgrenze.  Central- 
blatt  fiir  das  gesamte  Forstwesen.  87:  42-60. 

Turner,  H.;  Streule,  A.  1983.  Wurzelwachstum  iind 
Sprossentwicklung  junger  Koniferen  im  Klimastress 
der  alpinen  Waldgrenze,  mit  Berucksichtigung  von 
Mikroklima,  Photosynthese  und  Stoffproduktion. 
Irdning,  Austria:  Bundesanstalt  Gimipenstein:  617-635. 

Ulmer,  W.  1937.  Uber  den  Jahresgang  der  Frostharte 
einiger  immergriiner  Arten  der  alpinen  Stufe,  sowie  der 
Zirbe  und  Fichte.  Jahrbiicher  fur  wissenschaftliche 
Botanik.  84:  553-592. 


66 


A  PATTERN  OF  PINUS  PUMILA  SEED 
PRODUCTION  ECOLOGY  IN  THE 
MOUNTAINS  OF  CENTRAL 
KAMTCHATKA 

Peter  A,  Khomentovsky 


Abstract— This  paper  discusses  the  results  of  Pinus  pumila  cone 
production  studies  in  the  mountains  of  central  Kamtchatka.  Pi- 
nus pumila  was  found  to  have  a  very  high  seed  production  poten- 
tial when  not  overtopped  by  neighboring  trees.  When  shaded, 
few  or  no  cones  were  produced  on  P.  pumila.  Seed  production 
was  not  significantly  dependent  on  altitude  and  site  conditions. 
However,  elevations  above  1,300-1,400  m  were  unfavorable  for 
cone  survival,  mainly  because  of  low  temperature  and  phenologi- 
cal  delay.  Successful  seed  production  requires  snow  cover  to 
protect  ovulate  buds  on  low-lying  branches  from  freezing  in  the 
winter.  In  the  summer,  good  seed  production  is  related  to  high 
insolation,  wind  protection  on  cold  sites,  absence  of  shading,  and 
good  drainage  for  the  root  systems.  At  midelevations,  under  good 
seed  production  conditions  in  Kamtchatka,  about  110  kg  of  seeds 
per  hectare  will  be  produced  as  compared  to  only  1  kg  at  the  up- 
per elevational  limits.  Pinus  pumila  appears  to  have  a  high  evo- 
lutionary potential  that  reveals  itself  in  prominent  and  wide 
polymorphism  of  reproductive  organs. 


The  main  purpose  of  this  paper  is  to  give  a  brief  descrip- 
tion of  some  reproductive  characteristics  and  ecological 
features  of  typical  subalpine  dwarf  pine  (Pinus  pumila 
[Pall.]  Regel)  (Pinaceae,  Strobus)  (Pp)  for  many  regions 
of  its  area  ecotopes  of  the  central  Kamtchatka  Mountains 
in  northeastern  Asia  (fig.  1). 

Some  years  ago,  we  tried  to  identify  the  response  of 
seed  production  (reflected  in  cones,  seed,  and  nuclei  masses 
and  sizes)  to  some  of  the  most  evident  differences  of  Pp 
environments,  mainly  to  ecotopes,  altitudinal  position, 
and  their  location  in  regions  with  maritime  or  continental 
climate.  We  noticed  (Khomentovsky  and  Khomentovskaya 
1990)  that  there  is  no  strong  dependence  of  seed  param- 
eters and  cone  crop  on  the  ecotope,  elevation,  and  its  posi- 
tion in  the  relief,  Pp  was  supposed  to  have  some  ecotopic 
(macro-  or  microclimatic)  optimum  of  seed  production, 
which  could  be  indirectly  characterized  by  the  climate 
within  the  belt  of  600  to  900  m  above  sea  level  in  the  inte- 
rior continental  climate  regions  of  Kamtchatka  peninsula, 
and  widened  in  its  lower  limit  almost  to  sea  level  near  the 
coast. 

At  the  same  time  we  tried  to  develop  some  of  the  first 
descriptions  of  conophagous  insects  feeding  on  Pp  and  to 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Peter  A.  Khomentovsky  is  Head  of  the  Biogeocenology  Laboratory, 
Kamtchatka  Institute  of  Ecology  and  Nature  Management,  Prospect 
Rybakov,  19-A,  Petropavlovsk-Kamtchatsky  683024  Russia. 


define  their  role  in  cone  crop  size  and  quality  variation. 
We  found  (Khomentovsky  and  Efremova  1991)  that  the 
only  two  presently  known  insect  species  inhabiting  Pp 
cones  are  Cecidomyia  pumila  Mamaev  and  Efi-emova 
(Diptera,  Cecidomyidae)  and  Eupithecia  abietaria  Groeze 
(Lepidoptera,  Geometridae)  (the  taxonomy  of  the  last  spe- 
cies has  to  be  rechecked),  and  they  affect  crop  quality  very 
little.  Therefore,  this  damage  could  be  treated  as  negli- 
gible or  at  least  not  proven  to  be  important. 

These  preliminary  conclusions  forced  us  to  proceed  with 
a  more  detailed  investigation  of  Pp  seed  production.  Hav- 
ing no  possibilities  for  obtaining  direct  microclimatic  in- 
formation, we  decided  to  define  possible  principal  responses 
of  Pp  seed  production  to  elevation  and  ecotopical  features 
within  a  small  creek  basin  with  a  high  altitudinal  gradient. 

THE  RESEARCH  AREA 

The  research  area  is  situated  almost  in  the  geographi- 
cal center  of  the  Kamtchatka  (56  °N.,  158  °E.),  in  the 
Sredinnij  Mountain  Range  system,  the  main  system 
in  the  peninsula  (fig.  1).  According  to  the  botanical- 
geographical  regionalization  scheme  of  L.  F.  Kimitsin 
(1963),  this  territory  is  included  in  the  "mountain  above 
alpine  tundra  shrubby  region."  According  to  oar  scheme 
(Khomentovsky  and  others  1989),  it  lies  in  the  contact 
area  of  "Central  Kamtchatka  plainland-submountain 
province  of  conifer-stonebirch  forests"  and  "Sredinno- 
Western  middle  mountain  stone  birch-tundra  forest  prov- 
ince." The  area  covers  the  upper  part  of  the  conifer  forest 
vegetation  belt,  subalpine,  alpine,  and  above-alpine  belts 
in  Kamtchatka  (figs.  1  and  2). 

Climate  is  subcontinental,  with  300  to  400  mm  of  an- 
nual precipitation,  moderately  cold  and  snowy  winters, 
and  cool  summers  (monthly  average  air  temperature  in 
Esso  village  is  3  °C  in  May,  9.5  °C  in  June,  13  °C  in  July, 
12  °C  in  August,  and  6  °C  in  September.  The  permanent 
snow  cover  at  elevations  of  700  m  and  higher  begins  to 
accumulate  in  the  first  half  of  October  and  disappears 
mEiinly  in  May. 

Orographically,  the  area  can  be  called  "typical  middle 
altitudinal-mountainous"  (according  to  common  under- 
standing), which  for  Kamtchatka  actually  means  middle 
and  high  mountains  (due  to  the  compression  of  its  altitu- 
dinal vegetation  belts).  Average  height  of  the  mountains 
is  about  1,500  m.  Geological  and  geomorphological  struc- 
ture of  the  area  is  determined  by  volcanogenic  basalts  and 
andesits  (Ql,  Q2,  Q4),  and  by  quatenary  glacial,  fluvio- 
glacial  (Q3),  alluvial,  and  proluvial  (Q4)  deposits.  Glacial 
relief  is  conspicuous,  especially  in  the  upper  parts  of  river 


67 


I 


Figure  1 — General  description  of  the  research  area  in  Kamtchatka 
peninsula.  Top  left:  principal  scheme  of  vegetation  belts:  1 — subalpine, 
2 — boreal  forest,  3 — alpine,  4 — coastal  and  lowland.  Top  right:  main 
mountain  ridges  and  hydrological  net.  Bottom  left:  research  area,  within 
small  basin.  Dots  are  points  where  material  was  collected. 


more  or  less  permanent  (on  large  time  scale)  weak  or 
moderate  volcanic  ashfalls  impacts  (mainly  from  the  east- 
ern part  of  peninsula). 

Due  to  the  geographical  position,  mountain  relief,  and 
severe  climate  of  the  area,  vegetation  is  not  very  diverse 
and  exhibits  a  mosaic-like  spatial  pattern.  River  valleys 
and  lower  parts  of  their  tributaries  are  occupied  by  Popu- 
lus  suaveolens,  Chosenia  arbutifolia,  Larix  cajanderi, 


and  creek  basins.  Soils  are  typical  for  the  Kamtchatka 
Mountain  regions  classification  (I.  A.  Sokolov  1973),  with 
the  additions  by  N.  V.  Kazakov  (personal  communication) 
They  include  turf  illuvio-humus,  turf  illuvio-volcanic  ash, 
tundra  illuvio-humus,  tundra  illuvio-humus- volcanic  ash, 
and  turf  primitive.  Soil  genetics  reveal  themselves  in  the 
existence  of  specific  features  including  the  long-term  fi'eez- 
ing  period,  anaerobic  reactions  in  the  lower  horizons,  and 


68 


Figure  2 — A  fragment  of  the  vegetation  cover  in  the  upper  part  of  the  altitudinal  transect  (see  also  fig.  1). 


Betula  kamtschatica,  andB.  ermanii.  Two  dwarf  tree  spe- 
cies, Pinus  pumila  audi  Alnus  kamtschatica,  are  widely 
spread  along  the  whole  vegetation  profile.  They  occur  by 
groups  or  strips  along  the  moraine  tongues,  river  terraces, 
fluvioglacial  deposits,  and  watershed  ridges  and  slopes, 
and  form  the  upper  limit  of  woody  vegetation  (fig.  2). 

Plant  formations  in  the  middle  and  high  altitudes  are 
almost  the  same  (except  for  Populus  and  Chosenia),  but, 
moving  upward  along  the  narrowing  valleys  and  water- 
sheds, along  the  altitudinal  environmental  gradient  of 
gradually  decreasing  plant  life  conditions,  we  can  see 
the  distribution  pattern  of  the  formation  changing  from 
"macromosaic"  to  "micromosaic."  Tree  distribution  is 
more  strictly  related  to  soil  temperature  conditions  and 
drainage  gradients,  and  to  sites  sheltered  from  wind  and 
snow  abrasion.  This  is  true  even  for  such  ecologically 
flexible  dwarf  species  as  Pinus  pumila  and  Alnus 
kamtschatica. 

The  upper  elevational  limits  of  upright-grovnng  woody 
vegetation  {Larix  kamtchatica,  Betula  ermanii)  are  lo- 
cated between  900  to  1,000  m,  depending  on  site  condi- 
tions. Pinus  pumila  and  Alnus  kamtschatica  occur  up  to 
1,300  to  1,400  m  (rare  cl\mips).  Up  to  1,200  m  P.  pumila 
is  able  to  produce  seeds. 

MATERIALS  AND  METHODS 

Data  were  gathered  mostly  in  September  to  October  1990 
(with  some  additional  data  from  other  years)  at  the  field 
base  "Bolgit,"  in  the  vicinity  of  Esso  village  (Bistrinsky 
district  of  Kamtchatka  region),  in  the  "Tupikin  clyutch" 
creek  basin,  within  an  area  of  about  10  km^.  This  paper 
presents  the  material  from  one  slope  exposed  to  the  east- 
northeast  and  from  the  area  nearest  to  the  watershed  (up 
to  50  m  from  the  slope  edge),  within  an  altitudinal  range 
of  650  to  1,030  m.  Permanent  plots  on  an  existing  tran- 
sect and  some  additional  temporary  points  were  chosen 
for  the  cone  crop  measuring  and  sampling  (fig.  1). 

The  follovmig  information  was  collected  on  each  sample 
site  (table  1):  type  of  plant  commimity,  topographic 


location  and  ground  cover  of  Pp  clumps,  average  seasonal 
height  of  skeleton  branches,  general  moisture  conditions, 
shading  by  neighboring  upright  trees  or  climips. 

EXPLANATORY  COMMENTS 

First,  strictly  speaking,  we  cannot  consider  Pp  plants 
as  common  trees — they  are  not  standing  separately  and 
do  not  have  a  single  trunk;  in  most  cases  they  overlap 
each  other,  both  aboveground  and  in  imderground  parts. 
Also,  we  cannot  call  them  "clones"  because  we  do  not  know 
their  origin  and  level  of  genetic  relationships — which  is 
especially  important  in  the  case  of  seed  dispersal  by  ani- 
mals. The  best  term  would  be  "clump"  (we  are  grateful  to 
Dr.  Diana  Tomback  for  some  explanations  on  this  topic). 

Second,  the  single  trunk  (in  traditional  understanding) 
of  Pp  is  considered  a  creeping  tree  not  a  bush,  and  its 
length  is  only  several  centimeters  or  tens  of  centimeters. 
Branches  of  first  and  second  order  (I  prefer  to  call  them 
"trunk-branches")  exhibit  the  same  physiognomy,  bear 
similar  quantity  of  shoots,  and  form  multi trunk-like 
crovms.  In  the  case  of  syngenetic  origin  and  continuous 
dispersal,  Pp  plants  form  such  dense  cover  by  overlapping 
basal  parts  in  the  litter  and  root  systems  that  it  is  impos- 
sible to  identify  single  plants  and  thus  their  total  num- 
bers without  destroying  the  stand.  This  forced  us  to  use 
the  only  acceptable  method  of  measuring  cone  crops  or 
similar  parameters — instead  of  measuring  only  one  tree 
or  clump,  measure  by  square  unit  such  as  hectares  or 
square  meters.  Certainly,  the  stand  structure  and  charac- 
teristics of  groimd  cover  have  to  be  taken  into  account, 
especially  for  practical  taxation. 

Third,  it  is  also  impossible,  in  many  cases,  to  determine 
real  age  of  adult  Pp  plants.  They  produce  adventitious 
roots,  are  growing  apically,  and  are  gradually  decaying 
in  their  basal  parts  during  most  of  their  lifetime,  theoreti- 
cally endlessly.  We  have  to  recognize  that  we  really  can 
only  know  the  age  at  the  moment  of  measuring,  similar 
to  being  counted  in  young  trees,  and  not  the  real  one, 
which  is  usually  higher. 


69 


Table  1 — Short  description  of  the  ecotopes  and  Pinuspumila  clumps  being  analyzed.  The  age  of  clumps  in  these  analyses  lies  between  150 
and  260  years' 


Altitude, 

Pinus  pumila 

Way  of 
main  water 

Woodstand 

Pinus  pumila 

Shading 

m  asl 

Position  In  the  relief 

community  type 

supply 

composition 

H  avg,  cm 

SO,  % 

UTr 

Ppc 

1,030 

Upper  part  of  the  flat 
watershed,  shaded  from 
S  and  E  by  the  ridge 

Pumilae  pinetum 
carioso-hypnoso- 
ericosum,  with 
fragments  of  P.p. 
cladinosum  (CHE*C^ 

A+S 

lOPp 

100  (15-20) 

80  unev 

0 

1 

950 

Plateau,  opened  to  all 
directions,  watershed 
with  slight  slope  to 
NNE 

P.p.  purum,  with 
fragments  of 
P.p.  carioso- 
cladinosum  (P*CC^ 

A 

lOPp 

40  (10-15) 

40  unev 

0 

0 

900 

Flat  watershed  above 
the  creek  source, 
shaded  by  the  ridge 
from  S 

P.p.  hypnoso- 

carioso-ericosum 

(HCE) 

A+S 

lOPp+Lk 
(upper 
limit  of 
Lk) 

300 

40  unev 

1 

1 

810 

A  ridge  of  the  complex 
watershed,  slight 
slope  to  the  NE 

P.p.  carioso- 

hypnoso-ericosum 

(CHE) 

A 

lOPp+Lk 

150  (35-40) 

60  unev 

1 

2 

800 

Middle  part  of  the  E 
exposed  slope  of  wide 
creek  valley 

P.p.  hypnoso- 

carioso-ericosum 

(HE) 

S+A 

lOPp+Lk 

200 

80 

0 

1 

680 

Eastern  border  of  the 
watershed  ridge  with 
flattened  top 

P.p.  hypnoso- 

carioso-ericosum 

(HCE) 

A+S 

7Pp  3Lk 

300 

60  unev 

2 

1 

650 

Lower  part  of  E 
exposed  slope  in 
narrow  creek  valley 

P.p.  ericoso- 
sphagnosum 

8 

lOPp+Ak 

300  (40-45) 

100 

0 

3 

^ Pinus  pumila  community  type:  in  parentheses — abbreviation  for  fig.  8;  Way  of  main  water  supply:  A — atmospheric,  S — slope;  Woodstand  composition:  Pp — Pinus 
pumila,  Lk — Larix  cajanderi,  Ak — Alnus  kamtschatica  (dwarf  alder);  Pinus  pumila:  H  avg — average  height  (cm),  in  parentheses — rough  estimation  of  annual  shoot 
elongation  (mm);  SC% — surface  covering  by  the  clumps,  "unev" — uneven;  Shading:  Utr— from  upright  trees  of  different  species,  Ppc — from  Pinuspumila  neighbor- 
ing clumps;  0 — no  shading,  1 — slight,  2 — moderate,  3 — heavy  shading. 


In  each  point  for  data  and  sample  collecting  v^^e  used 
3  to  5  small  randomized  plots  of  different  size — from  1  to 
15  m^.  The  following  parameters  were  countered  in  place 
(table  2):  number  of  skeleton  branches  ("trunk-branches"), 
number  of  germinating  shoots  on  each  skeleton  branch; 
number  of  current  year  crop  cones  (the  second  year  of 
cone  development)  and  number  of  next  year  crop  cones 
(female  buds  of  the  first  year)  on  each  germinating  shoot 
and  each  skeleton  branch;  and  among  the  total  quantity  of 
each  year's  cones  the  number  of  cones  damaged  by  insects 
iCecidomyia  pumila — damage  in  the  first  year  of  cone  de- 
velopment, Eupithecia  abietaria — damage  in  the  second 
year  of  cone  development),  and  by  the  nutcracker  {Nuci- 
fraga  caryocatactes  kamtschatkensis) . 

At  each  site,  10  to  100  cones  of  the  current  year's  crop 
were  collected  and  measured  before  they  dried  (cone 
length  and  cone  diameter).  Later,  after  air  drying  during 
some  weeks  or  months,  cones  and  seeds  were  measured 
by  16  other  parameters,  six  of  which  are  discussed  in  this 
paper:  cone  mass;  1,000  seeds  mass;  1,000  nuclei  mass 
(both  by  measuring  a  number  of  samples  with  50  seeds  in 
each);  nimiber  of  seed  scales  in  the  cone;  among  the  total 


number  of  seed  scales,  the  number  of  scales  not  contain- 
ing seeds  under  them;  and  total  seed  quantity  in  the  cone. 

RESULTS  AND  BRIEF  COMMENTS 

•  The  structure  of  the  cone-bearing  portion  of  the 
crown  and  cone  crop  estimates  (tables  1  and  2,  figs.  1-6): 

As  a  whole,  the  number  of  skeleton  branches  and  germi- 
nating shoots  per  hectare  varied  very  little  within  alti- 
tudes of  650-800  m  and  definitely  increased  upward.  This 
could  be  explained  by  increasing  isolation.  Most  likely  it 
was  the  reason  for  the  especially  increasing  number  of 
germinating  shoots  (two  to  three  times  higher  than  skel- 
eton branches)  and,  correspondingly,  the  nvimbers  of 
current-year  and  next-year  cones.  The  last  varied  more 
than  the  first  because  some  of  them  vdll  inevitably  die  in 
the  natural  selection  process  during  the  maturation  time 
in  the  coming  year. 

The  number  of  shoots  with  cones  per  skeleton  branch 
(fig.  4)  changed  opposite  to  that  of  the  number  of  skeleton 
branches  per  1  hectare  (we  noticed  the  increased  number 


70 


of  shoots,  including  nongerminating).  Even  a  slight  plant 
shading  immediately  resulted  in  a  change  in  the  number 
of  shoots  (sites  900  m  and  1,030  m). 

The  quantity  of  current  and  next-year  crop  cones  per 
one  germinating  shoot  had  specific  dispersal  features: 
at  each  site,  especially  in  the  lower,  shaded  part  of  the 
elevation  profile,  cones  of  any  one  year  were  overwhelm- 
ing. This  migration  of  intensive  seed-producing  centers 
from  year  to  year  supports  our  previous  conclusion 


(Khomentovsky  and  Efremova  1991)  about  the  existence 
of  local  seed  production  mosaics,  which  provides  a  con- 
tinuous supply  of  seeds,  important  not  only  for  guaran- 
teed reproduction  and  microevolutionary  diversity  for  Pp, 
but  also  for  all  zooconsumers  such  as  nutcracker,  other 
birds,  and  mammals. 

•  Cones  and  seeds  mass  and  size  variation  (tables  1,  3; 
figs.  7  and  8): 


Table  2 — Characteristics  of  germinating  crown  parts  and  cone  crop  outcome  of  Pinus  pumila  in  central  Kamtcliatka  Mountains^ 


nn 

Parameter 

Altitude  (m  above  sea  level) 

650 

680 

800 

810 

900 

950 

1,030 

1 

No.  of  skeleton  branches/ha 

1,200 

4,167 

2,240 

21,111 

11,000 

30,000 

10,333 

2 

No.  of  germinating  shoots/ha 

1,733 

7,917 

3,200 

47,778 

57,000 

86,667 

33,000 

3 

No.  of  this  year  crop  cones/ha 

1,733 

4,375 

3,136 

2,389 

46,170 

56,334 

28,000 

4 

No.  of  next  year  crop  cones/ha 

260 

0 

160 

51,122 

19,950 

143,000 

24,550 

5 

No  of  aerminatina  shoots/skel  bran 

1.44 

1.90 

1.43 

2.26 

5.19 

2.89 

3.00 

6 

No  of  thi^  vppir  oroo  conp^/<>kpl  hri^n 

1.44 

2.00 

1.40 

0.11 

4.20 

1.88 

2.56 

7 

No  of  next  vear  croo  cones  /skel  bran 

0.22 

0 

A  A7 

O  AO 

1  Md. 

4.77 

2.54 

o 
O 

No.  of  this  year  crop  cones/germ,  shoot 

1 .00 

H  AC 
1  .00 

0.98 

0.05 

0.81 

A 

A  7ft 

9 

No.  of  next  year  crop  cones/germ,  shoot 

0.15 

0 

0.05 

1.07 

0.35 

1.65 

0.86 

10 

Percentage  of  skeleton  branches  with 

various  numbers  of  germinating  shoots: 

with  one  shoot 

67 

30 

64 

58 

28 

22 

32 

with  two  shoots 

22 

CA 

50 

29 

16 

18 

J4 

with  thrfifi  shoots 

will  1   11  II           Ol  Iwwlw 

11 

20 

7 

16 

0 

22 

8 

with  four  shoots 

0 

0 

0 

0 

9 

11 

4 

with  fiv/o  chnr^tQ 
Will  i  live?  Ol  IwUlO 

0 

0 

0 

5 

9 

0 

8 

with  qIy  QhnntQ 

Willi  OlA  Ol  lUUlO 

0 

0 

A 
0 

A 
0 

A 
0 

0 

2 

with  seven  shoots 

0 

0 

0 

0 

0 

0 

5 

with  eight  shoots 

0 

0 

0 

0 

9 

11 

5 

with  nine  shoots 

0 

0 

0 

0 

g 

0 

0 

with  ten  shoots 

0 

0 

0 

0 

0 

0 

2 

with  eleven  shoots 

0 

0 

0 

0 

0 

0 

0 

with  twelve  shoots 

0 

0 

0 

5 

0 

0 

0 

with  thirteen  shoots 

0 

0 

0 

0 

Q 

0 

1 

1 1 

r  ciccf iictgc  ui  gcniiiriaurig  biiouis  Dcaiirig 

various  number  of  this  year  cones  (T) 

and  next  year  cones  (N): 

0N-1T 

81 

100 

93 

86 

63 

8 

32 

ON -21 

7 

0 

2 

8 

4 

0 

6 

Cumulative  percent  of  shoots  with 

this  year's  crop  cones  only 

88 

100 

95 

94 

67 

8 

38 

1N-0T 

4 

0 

0 

3 

25 

15 

30 

2N-0T 

4 

0 

5 

0 

2 

11 

7 

3N-0T 

0 

0 

0 

0 

0 

11 

0 

Cumulative  percent  of  shoots  with 

this  year's  crop  cones  only 

8 

0 

5 

3 

27 

37 

37 

1N-1T 

4 

0 

0 

3 

3 

23 

10 

1N  -2T 

0 

0 

0 

0 

3 

4 

2 

2N  -  1T 

0 

0 

0 

0 

0 

15 

6 

2N-2T 

0 

0 

0 

0 

0 

5 

5 

3N-1T 

0 

0 

0 

0 

0 

8 

1 

4N  -  1T 

0 

0 

0 

0 

0 

0 

1 

Cumulative  percent  of  shoots  with 

cones  of  both  years 

4 

0 

0 

3 

6 

55 

25 

'Only  the  area  covered  with  P'mus  pumila  clumps  is  taken  into  account  (par.  1  -4),  not  the  whole  territory  of  the  site;  number  of  next-year  cones  can  decrease 
during  the  year  due  to  natural  selection  processes. 


71 


Table  3 — Some  characteristics  of  Pinus  pumila  cones  and  seeds  in  various  elevations  in  central 
Kamtchatka  Mountains  (within  research  area)' 


Regression  on  altitude  Correlation  Average  value 


CL 

=  54.71 



0.020  A 

R 

=  -0.73 

AVG  (cl)  = 

43 

+/- 

8  (25-62) 

CD 

=  33.68 

- 

0.010  A 

R 

=  -0.79 

AVG  (cd)  = 

27 

+/- 

2(18-37) 

CM 

=  7.81 

0.030  A 

R 

=  -0.41 

AVG  (cm)  = 

7 

+/- 

1  (4-10) 

SSN 

=  32.30 

+ 

0.010  A 

R 

=  0.53 

AVG(ssn)  = 

39 

+/- 

5  (26-52) 

SWS 

=  9.06 

+ 

0.004  A 

R 

=  0.33 

AVG  (SWS)  = 

11 

+/- 

3  (5-18) 

SQT 

=  28.50 

+ 

0.003  A 

R 

=  0.76 

AVG  (sqt)  = 

45 

+/- 

5  (26-68) 

SMP 

=  46.71 

+ 

0.002  A 

R 

=  -0.11 

AVG(smp)  = 

45 

+/- 

3  (32-55) 

SMTH 

=  101.67 

0.020  A 

R 

=  -0.45 

AVG  (smth)  = 

84 

+/- 

8  (52-116) 

NMTH 

=  56.49 

0.020  A 

R 

=  -0.49 

AVG  (nmth)  = 

43 

+/- 

4  (27-68) 

'CL — cone  length,  mm;  CD — cone  diameter,  mm;  CM — cone  mass,  g;  A — altitude,  m  above  sea  level;  SSN — seed 
scale  number;  SWS— number  of  scales  without  seeds  (within  SSN);  SQT — total  number  of  seeds  in  the  cone;  SMP — 
percentage  of  seeds  mass  in  cone  mass;  SMTH — mass  of  1 ,000  seeds,  g;  NMTH — mass  of  1 ,000  nuclei,  g.  In  paren- 
theses at  the  right  side — minimum  and  maximum  values. 


Mass  variation  (here  and  later  we  speak  about  current- 
year  crop  cones  only)  is  more  evident  in  the  upper  part  of 
the  elevation  profile,  in  micromosaic  sites  with  increasing 
abiotic  environmental  pressure.  In  sufficiently  insolated 
(sunny)  and  wind-protected  sites  cone  mass  is  increasing, 
in  shaded  sites  or  windy  plateaus — decreasing.  At  the  up- 
per elevation  levels  of  Pp  distribution  (1,200-1,300  m)  we 
observed  natural  abrupt  decreases  of  all  seed-producing 
parameters.  It  is  remarkable  that  in  general  cones  and 
seeds  masses  are  not  too  closely  related  to  definite  altitu- 
dinal  or  site  positions  of  the  tree,  even  in  cases  of  evi- 
dently unfavorable  envirormients  (site  950  m). 

Self  regulation  of  seed  production  is  obvious  also:  along 
the  elevation  gradient,  number  of  seed  scales  in  the  cone 


(with  and  without  seeds)  and  total  number  of  seeds  in  the 
cone  slightly  increases  as  well,  and  at  the  same  time  seed 
and  nuclei  mass  is  unchanging  or  slightly  decreasing.  It 
means  that  seeds  become  smaller  but  grow  in  quantity, 
keeping  the  same  reproduction  potential  and  having  more 
protection  from  environmental  impacts. 

Cone  size  variation,  most  evident  in  their  length,  corre- 
sponds vrith  mass  variation  in  relation  to  site  position  in 
the  landscape.  Here,  as  well  as  in  the  previous  case,  the 
main  impact  factor  is  not  elevation  but  landscape  struc- 
ture of  the  ecotope  (slopes  in  wide  or  narrow  valley,  pla- 
teau, watershed  ridge,  etc.).  Biocenotic  structure  is  not 
too  important  in  this  case,  at  least  at  the  forest  type  level. 
All  ecotopes,  excluding  the  950-m  site  (opened  plateau), 


160 


140 


§  120 
O 


£  100 

(0 

u 

X  80 

CL 

O  60 

n 
E 

3 

Z  40 


20 


i 

\ 

\ 

i 
i 

\ 

\ 

— \  

— i-/- 

\ 

.  \ 

/ 
/I 

\  \ 
\\ 

 V 

\  1 

•  y 

>' 
/ 

f 

\l 
1  • 

1  . 

1  / 

1  ^ 

 skeleton  branches 

 germinating  shoots 

—  current  year  cones 
 next  year  cones 


650    680    800    810    900    950  1.030 

Altitude  (m  as!) 

Figure  3 — Principal  structure  of  P'mus  pumila  crowns  at 
various  altitudes  and  estimated  cone  crop  per  hectare. 


72 


6 


germ,  shoots/sk.  br. 


650        680        800         810        900         950  1,030 


Altitude  (m  asl) 

Figure  4 — Structure  of  Pinus  pumila  crown  germinating 
part  (number  of  germinating  shoots  and  estimated  cone 
crop  per  skeleton  branch). 


are  occupied  by  the  same  group  of  forest  types — Pumilae 
pinetum  carioso-hypnoso-ericosum,  and  its  variations. 
Generally  speaking,  cones  size,  as  well  as  cones  mass, 
which  has  good  correlation  with  them  (r  =  0.7-0.8),  vary 
within  limits  already  known  for  the  species. 

•  Damage  caused  by  insects  and  birds  (fig.  9): 

This  was  investigated  in  some  detail  previously 
(Khomentovsky  and  Efremova  1991)  and  is  not  likely 


to  effect  any  substantial  change  in  average  cone  sizes:  it 
does  not  affect  cone  diameter  and  leads  to  cone  length  de- 
creases up  to  3-9  percent  (Cecidomyia  pumila).  However, 
in  some  cases  this  insect  species  damage  results  in  under- 
development of  up  to  20  percent  of  the  seeds,  but  this  im- 
pact usually  does  not  stop  the  development  of  the  rest  of 
the  seeds  in  the  cone  and  is  compensated  at  the  popula- 
tion level  by  the  abundant  production  of  seeds. 


73 


100 


cones  of  both  years 
next  year  cones  only 
current  year  cones  only 


650    680  800    810    900  950  1,030 

Altitude  (m  asl) 

Figure  6 — Distribution  of  current-year  and  next-year  cone  crop  on 
germinating  shoots. 


The  level  of  damage  caused  by  Cecidomyia  pumila, 
clearly  noticeable  visually  by  the  cone  curving,  is  usually 
rather  high  everywhere  but  it  varies  with  different  site, 
environmental,  and  weather  conditions,  to  which  insects 
are  more  sensitive.  In  our  case  it  was  changing  from  mod- 
erate damage  of  25-50  percent  on  the  650-m  site  (slope 
in  the  narrow  valley),  the  800-m  site  (slope  in  the  wide 


valley),  and  the  900-m  site  (wind-protected  watershed  at 
the  upper  limit  of  Larix)  to  a  maximum  of  100  percent  on 
the  810-m  site  (watershed  ridge)  and  a  minimal  8  percent 
on  the  1,030-m  site  (weakly  wind  protected  and  too  cold 
for  insects  ecotope). 

Confirmation  of  the  negligible  character  and  amount  of 
insect  damage  is  also  seen  in  the  level  of  seed  extraction 


o 


o 
m 

w 

"O 

<u 
a> 
0) 
o 
in 

V) 

<a 
c 
o 
O 


650 


800 


900 


1,030 


O 

o 

CM 
II 

o 

(/) 


Altitude  (m  asl) 


cone  mass 
SM  std 


  CM  std 

  50  nuclei  mass 


50  seeds  mass 
NMstd 


Figure  7 — Variations  in  Pinus  purpila  cones,  seeds,  and  nuclei  mass. 
SL — slopes,  P — Plateau. 


74 


650  680  800  810  900  950  1,030 


Altitude  (m  asl) 

Figure  8 — Variations  in  Pinus  pumila  cone  size.  1 — forest  type  (see  table  1). 


by  the  nutcracker,  which  collects  only  full,  matured  seeds 
(fig.  9). 

DISCUSSION  AND  CONCLUSIONS 

Data  analysis  results,  combined  with  some  previous 
material,  allow  us  to  make  some  conclusions  about  Pinus 
pumila  seed  production  in  Kamtchatka,  keeping  in  mind 
the  fragmentary  nature  of  our  study,  even  though  it  was 
done  in  typical  environmental  conditions. 

1.  The  main  conclusion  is  that  P.  pumila  seed  produc- 
tion has  such  a  high  potential  in  time  and  space  that  we 
can  speak  about  the  environmental  (abiotic  and  natural 
biotic)  impact  on  it  in  the  frame  of  modification  only  but 
not  in  the  frame  of  regulation.  Environmental  stresses 
usually  cannot  cause  a  crop  crisis — trees  have  compensat- 
ing mechanisms  both  at  the  organismic  and  population 


level,  and  one  of  them  is  the  microhabitat  mosaics  of  an- 
nual seed  production. 

2.  Pinus  pumila  seed  production  is  not  (in  Kamtchatka, 
at  least)  significantly  dependent  on  altitude  and  site  con- 
ditions. Landscape  conditions  (above  facial  level)  are  the 
most  important.  However,  this  is  true  only  when  trees 
are  not  overshadowed  by  neighboring  upright  trees;  that 
is,  when  Pp  forms  an  independent  vegetational  belt  in 
more  or  less  solid  segments — in  plain  or  mountain  tundra, 
upper  forest-tundra  ecotone,  on  the  seashore  dunes,  in 
wide  river  valleys,  etc.  Pp  often  grows  as  the  lower  canopy 
of  more  or  less  dense  Larix  forests  (more  than  30  percent 
of  Larix  cover),  which  have  the  same  age  and  the  same 
syngenetic  origin  (very  often  in  pyrogenetic  ecotopes). 
These  shaded  dwarf  pine  usually  do  not  produce  seeds 
and  develop  vegetatively  only. 

3.  The  utmost  upper  limits  of  distribution  (1,300-1,400  m 
above  sea  level)  are  unfavorable  for  Pp  seed  production. 


650 


680 


800  810  900 

Altitude  (m  asl) 


950 


1,030 


Ea+Nc 


□ 


Cp 


Ea+Cp+Nc 


Nc 


Figure  9 — Percentage  of  cones  damaged  by  insects  and  birds.  Cp — Cecidomyia 
pumila;  Ea — Eupithecia  abietaria;  Nc — Nucifraga  caryocatactes. 


75 


mainly  because  of  low  temperature  and  phenological  de- 
lay. But  even  here,  some  cones  can  be  found,  and  some- 
times the  crop  can  be  abundant  in  wind-protected  habi- 
tats. Phenological  delay  often  prevents  the  nutcracker 
from  collecting  seeds  here.  These  seeds,  having  been  dis- 
persed by  birds  from  lower  sites,  and  having  more  stable 
genotypes,  provide  a  better  chance  for  successful  refores- 
tation in  new  areas.  At  the  same  time,  dwarf  pine  regen- 
eration at  high  altitudes  is  more  or  less  provided  by  the 
remaining  abundant  seeds.  Besides  this,  numerous  zoo- 
consorts  (mainly  micromammals)  will  have  a  good  food 
supply  most  valuable  in  extreme  conditions. 

The  same  unfavorable  conditions  for  seed  production 
can  be  found  in  the  Pp  belt  at  the  seashore  (Pacific 
Kamtchatka  coast),  on  the  dunes  nearest  to  the  water. 
But  here,  as  well  as  in  the  mountains,  trees  utilize  each 
opporttmity  for  maximal  realization  of  their  reproduction 
ability.  Pp  seed  production  on  the  second  to  third  dune 
of  the  coast  belt  is  equal  in  cone  size  and  mass  parameters 
to  those  at  the  900-  to  1,000-m  altitude  and  in  seed  and 
nuclei  parameters  to  those  at  the  650-  to  750-m  altitude 
in  the  continental  climate  portion  of  the  peninsula. 

Conditions  moderately  favorable  for  Pp  seed  production 
can  be  found  in  plain  and  submoimtain  (foothill)  places, 
for  instance  on  gravel-sandy  fluvioglacial  or  proluvial 
deposits,  where  abundant  cone  crops  sporadically  occur. 

The  most  productive  seed  zone  lies  in  the  vegetation 
belt  within  the  600-900  m  above  sea  level  area.  Here, 
cone  crops  are  mostly  permanent  and  stable  from  year 
to  year. 

4.  Successful  seed  production  of  Pp  needs  the  following 
conditions:  in  winter — sufficient  snow  cover  to  protect  buds 
of  low-ljdng  trees  from  freezing;  in  summer — high  insola- 
tion, good  wind  protection  on  cold  sites,  absence  of  shad- 
ing, and  good  drainage  for  the  root  systems.  The  main 
adaptive  feature  of  Pp  is  its  ability  to  lay  close  to  the 
ground  in  the  autumn.  This  process  is  initiated  by  low 
temperature  in  the  autumn  before  the  snow  cover.  This 
unique  feature  provides  good  chances  of  survival  under 
very  extreme  conditions.  Strict  dependence  of  Pp  on  snow 
cover  depth  is  proved  by  coincidence  of  its  area  border  with 
the  snow  depth  isoline  of  40  cm  and  more  (Lukitcheva 
1964;  Tstcherbakova  1964). 

5.  Special  research  and  preliminary  work,  including 
mapping  of  various  productivity  sites,  is  needed  for  the 
practical  (silvicultural)  estimation  of  current  or  future 
cone  crops  of  Pp.  It  is  possible  to  show,  as  an  example, 
the  scope  of  crop  diversity.  In  the  middle  elevation  site 
(700-800  m),  with  5  g  of  seed  mass  in  the  cone,  80  percent 
of  germinating  shoots,  and  80  percent  of  Pp  covering  the 
ground,  we  can  get  about  110  kg  of  seeds  per  hectare.  At 
the  upper  limit  of  seed  production,  with  2  g  of  seed  mass 
in  the  cone,  30  percent  of  germinating  shoots,  10  percent 
of  covering,  we  can  get  about  1  kg  of  seeds  per  hectare. 
High  variability  of  cone  crop  is  illustrated  in  table  1  data: 
taking  into  account  only  pure  stands  of  Pp  (without  con- 
sidering its  real  distribution  on  the  surface)  we  can  see 
that  the  crop  size  on  the  650-m  site  is  about  6.5  kg/ha, 
and  about  212  kg/ha  on  the  950-m  site. 

6.  Starting  from  the  elevation  that  corresponds  to  the 
upper  limit  of  upright  trees  distribution  (900  to  1,000  m) 


in  Kamtchatka,  Pp  produces  an  increasing  number  of 
shoots.  This  can  be  interpreted  as  a  organismic  compen- 
sation for  the  increasing  severity  of  abiotic  environmental 
conditions.  Photosynthetic  ability  increases,  seed  produc- 
tion potential,  as  we  saw,  remains  at  the  same  high  level, 
and  seed  protection  (thickness  of  seed  walls)  becomes 
stronger.  All  this  takes  place  mostly  in  the  subalpine  belt 
and  proves  Pp  belongs  mainly  to  this  type  of  vegetation. 
Coincidence  of  these  activated  processes  with  the  upper 
limit  of  upright  trees  distribution  allows  us  to  hypothe- 
size the  existence  of  some  kind  of  temperature  threshold- 
trigger.  It  could  be  the  sum  of  effective  temperatures, 
above  which  the  development  of  upright  tree  forms  is  pos- 
sible, and  below  which  only  prostrate  forms  develop. 

7.  We  agree  with  the  opinion  that  Pp,  as  well  as  other 
stone  pines,  has  a  high  evolutionary  potential,  revealing 
itself  in  prominent  and  wide  polymorphism  of  reproduc- 
tive organs  and  their  functional  flexibility  (Pravdin  and 
Iroshnikov  1982).  This  is  well  illustrated  by  our  work 
and  other  sources.  Ecotypical  features  are  also  evident, 
accompanied  by  stability  of  the  main  seed-production 
parameters.  In  fact  almost  all  of  this  is  similar  to  that 
which  is  known  from  other  Pp  regions  (Bobrinev  and 
Rylkov  1984;  Efremova  and  IvHev  1972;  Kapper  1954; 
Khomentovsky  and  Khomentovskaya  1990;  Krylov  and 
others  1983;  Rush  1974). 

8.  Pp  seed  production  potential  exceeds  the  possibilities 
of  its  realization  in  our  research  area,  and  most  likely,  in 
the  Kamtchatka  peninsula  as  a  whole.  Ecotopical  diver- 
sity in  Kamtchatka  is  evidently  only  a  part  of  the  whole 
scope  of  acceptable  conditions  for  the  species  (here  we 
can  admit  that  low  temperatures  affect  vegetative  growth 
three  to  four  times  more  than  they  affect  generative  or- 
gans variability).  Therefore,  we  should  look  for  possible 
phylogenetic  roots  of  this  species  in  the  regions  with  a 
wider  geoecological  spectrum  of  sites. 

9.  Our  observations  (partly  presented  in  this  paper)  al- 
low us  to  speculate  that  Pp  is  a  middle-altitudinal  species, 
with  more  preferable  growth  conditions  in  the  moderately 
continental  climate  (in  relatively  humid  mountains)  than 
at  the  seashore.  But,  regarding  the  whole  area,  the  defi- 
nite and  necessary  influence  of  Pacific  wet  air  masses  re- 
sults in  considerable  snow  cover. 

Combining  this  conclusion  with  some  known  hypothe- 
ses of  Pp  evolution  (Sotchava  and  Lukitcheva  1953; 
Tikhomirov  1949;  etc.)  we  think  that  this  dwarf  tree  spe- 
cies most  likely  appeared  in  Tertiary  times  in  some  cli- 
matically subcontinental  mountain  regions  of  Angarida 
(northeastern  Asia).  In  Kamtchatka,  pollen  data  show 
Pp  has  been  here  for  1-1.5  million  years  (Malaeva  1967; 
Tchelebaeva  and  others  1974).  During  Pleistocene  inter- 
glacial  periods,  and  in  the  early  middle  Holocene,  Pp 
occupied  periglacial  zones  and  other  places  unacceptable 
for  more-thermophil  tree  species.  The  highest  incidence 
of  polyembriony  and  underdevelopment  of  seeds  in 
Kamtchatka  support  the  possible  centers  of  origin  in  the 
moderate  continental  climates  when  compared  to  Magadan 
(northeastern  continental  part  of  Asia)  and  Buryatia 
(southern  part  of  eastern  Siberia)  (Iroshnikov  1972;  Rush 
1974).  These  features  can  be  considered  a  reaction  to  less 
favorable  conditions  of  the  secondary  environments. 


76 


including  Kamtchatka  peninsula.  According  to  A.  I. 
Iroshnikov  (1972),  Pp  is  one  of  the  most  well-adapted 
species  to  the  continental  climate  of  northeastern  Asia. 

REFERENCES 

Bobrinev,  V.  P.;  Rylkov,  V.  F.  1984.  Seed  production  of 
dwarf  stone  pine.  Inform,  bull,  of  Tchitinsky  Informa- 
tion Center,  #76-84.  Tchita.  5  p.  [In  Russian]. 

Efremova,  L.  S.;  Ivliev,  L.  A.  1972.  On  dwarf  stone  pine 
seed  producing  at  Kamtchatka.  In:  Utilization  and  re- 
producing of  forest  resources  in  the  Far  East.  Abstr.  of 
the  all-imion  conference  Khabarovsk.  Part  2:  158-159. 
[In  Russian]. 

Iroshnikov,  A.  I.  1972.  Seed  producing  and  seed  quality 
of  conifers  in  northern  and  mountainous  regions  of 
Siberia.  In:  Seed  producing  of  Siberian  conifers  (collec- 
tion of  papers).  Novosibirsk:  98-117.  [In  Russian]. 

Kapper,  O.  A.  1954.  Conifers  (forestry  characteristics). 
IVloscow-Leningrad.  304  p.  [In  Russian]. 

Khomentovsky,  P.  A.;  Efremova,  L.  S.  1991.  Seed  prod- 
uction and  cone-feeding  insects  oiPinus  pumila  at 
Kamtchatka  peninsula:  aspects  of  coexistence.  In: 
Baranchikov,  Y.  N.;  Mattson,  W.  J.;  Hain,  F.  P.;  Payne, 
T.  L.,  eds.  Forest  insects  guilds:  patterns  of  interaction 
with  host  trees.  U.S.  Department  of  Agriculture,  Forest 
Service,  Gen.  Tech.  Rep.  NE-153:  316-320. 

Khomentovsky,  P.  A.;  Kazakov,  N.  V.;  Tchernyagina,  O.  A. 
1989.  Tundra-forest  zone  at  Kamtchatka:  problems  of 
conservation  and  management.  In:  Problems  of  nature 
management  in  taiga  zone  (collection  of  papers). 
Irkutsk:  30-46.  [In  Russian]. 

Khomentovsky,  P.  A.;  Khomentovskaya,  I.  G.  1990.  Geo- 
graphical variability  of  Pinus  pumila  seed  producing  at 
Kamchatka.  In:  Questions  of  Kamchatka  geography. 
Petropavlovsk-Kamchatsky.  No.  10:  47-55.  [In  Russian]. 


Krylov,  G.  V.;  Talantsev,  N.  K;  Kozakova,  N.  F.  1983. 
Siberian  pine  (Pinus  sibirica  Du  Tour).  Moscow.  216  p. 
[In  Russian]. 

Kunitsin,  L.  F.  1963.  A  pattern  of  Kamtchatka  nature 
regionalization.  In:  Nature  conditions  and  regionaliza- 
tion  of  Kamtchatka  district.  Moscow:  7-26.  [In  Russian]. 

Lukitcheva,  A.  N.  1964.  Distribution  areas  of  Asia  plants 
(a  map).  In:  World  geographical  atlas.  Moscow.  112  p. 

Malaeva,  E.  M.  1967.  Kamtchatka  vegetation  develop- 
ment in  Pliocene-Pleistocene.  In:  Development  of  Sibe- 
rian and  far-eastern  vegetation  in  Quaternary  (collec- 
tion of  papers).  Moscow:  78-170.  [In  Russian]. 

Pravdin,  L.  F.;  Iroshnikov,  A.  I.  1982.  Genetics  of  Pinus 
sibirica  Du  Tour,  P.  koraiensis  Sieb.  et  Zucc.  and  P. 
pumila  Regel.  Annales  Forestales.  9(3):  79-123. 

Rush,  V.  A.  1974.  Biochemical  characteristics  of  five- 
needle  pines  seeds.  In:  Biology  of  conifer  seed  repro- 
duction in  western  Siberia  (collection  of  papers). 
Novosibirsk:  180-184.  [In  Russian]. 

Sokolov,  I.  A.  1973.  Volcanism  and  soils  formation 
(with  Kamtchatka  as  an  example).  Moscow.  224  p. 
[In  Russian]. 

Sotchava,  V.  B.;  Lukitcheva,  A.  N.  1953.  On  geography 
of  dwarf  stone  pine.  Reports  of  the  USSR  Academy  of 
Sciences.  90(6):  1163-1166.  [In  Russian]. 

Tchelebaeva,  A.  I.;  Shantser,  A.  E.;  Egorova,  I.  A.;  Lupikina,  E. 
G.  1974.  Cenozoic  deposits  of  the  Kuril-Kamtchatka 
area.  In:  Kamtchatka,  Kuril  and  Commander  islands 
(series  of  History  of  the  relief  development.)  Moscow: 
31-57.  [In  Russian]. 

Tikhomirov,  B.  A.  1949.  Dwarf  stone  pine  {Pinus  pumila 
(Pall.)  Rgl),  its  biology  and  utilization.  MOIP  series, 
botan.  division,  Moscow.  14(6):  106  p.  [In  Russian]. 

Tstcherbakova,  E.  J.  1964.  An  average  of  maximal  decade 
snow  depths  (a  map).  In:  World  geographical  atlas. 
Moscow.  220  p. 


77 


NEEDLE  LONGEVITY  AND 
PHOTOSYNTHETIC  PERFORMANCE  IN 
CEMBRAN  PINE  AND  NORWAY  SPRUCE 
GROWING  ON  THE  NORTH-  AND  EAST- 
FACING  SLOPES  AT  THE  TIMBERLINE 
OF  STILLBERG  IN  THE  SWISS  ALPS 

Takayoshi  Koike 
Rudolf  Hasler 
Hans  Item 


Abstract — The  number  of  shoots  in  each  age  class  and  the  pho- 
tosynthesis of  the  shoots  were  measured  on  cembran  pine  {Pinus 
cembra)  and  Norway  spruce  (Picea  abies)  growing  on  north-  and 
east-facing  slopes  at  2,185  m  above  sea  level.  Needle  longevity 
on  the  north-facing  slope  exceeded  needle  longevity  on  the  east- 
facing  slope  by  1  to  3  years  in  both  species.  A  negative  correla- 
tion between  maximum  photosynthesis  and  needle  lifespan  was 
foimd.  However,  there  was  no  physiological  difference  in  the 
photosynthesis  of  trees  on  the  respective  slopes. 


In  the  Swiss  Alps,  cembran  pine  (Pinus  cembra)  grov^rs 
around  timberline  level  and  Nonvay  spruce  (Picea  abies) 
from  valleys  to  mountainsides  (Kuoch  and  Amiet  1970). 
At  the  subalpine  timberline,  the  growth  and  biomass  in- 
crease of  trees  are  strongly  influenced  by  the  microsite 
performance  and  microclimate  (Livingston  and  Black 
1987;  Schonenberger  and  Frey  1988;  Tvu-ner  and  others 
1982).  Using  the  monitoring  experiment  on  photosyn- 
thetic  production  as  a  basis,  the  biomass  increase  of 
mountain  pine  (Pinus  mugo)  growdng  at  the  timberline 
was  found  to  be  regulated  by  the  soil  temperature  and  the 
net  radiation  (Hasler  1982).  However,  physiological  char- 
acteristics of  trees  were  not  revealed.  Is  the  turnover  rate 
of  needles  related  to  the  tree  size?  What  about  the  physi- 
ological adaption  of  needles  to  the  microclimate? 

In  this  study,  these  questions,  needle  lifespan  and  the 
age-related  photosynthesis  of  cembran  pine  planted  on 
north-  and  east-facing  slopes,  were  measured.  We  present 
the  physiological  parameters  of  needle  photosynthesis. 
Anatomical  traits  of  needles  were  investigated  with  refer- 
ence to  the  estimation  of  the  intercellular  CO^  concentra- 
tions (C.)  (Terashima  1992).  We  discuss  the  trade-off  rela- 
tion between  the  needle  lifespan  and  photosynthetic  rates 
influenced  by  the  microclimate. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Takayoshi  Koike,  Rudolf  Hasler,  and  Hans  Item  are  members  of  the 
"Ecophysiology"  group,  Swiss  Federal  Institute  for  Forest,  Snow  and  Land- 
scape Research,  CH-8903  Birmensdorf,  Switzerland.  Present  address  of 
T.  Koike:  Forest  Products  Research  Institute,  Sapporo  062,  Japan. 


MATERIALS  AND  METHODS 

At  the  study  site  (Stillberg  in  the  Swiss  Alps),  on  the 
north-facing  slope,  the  yearly  net  radiation  surplus  was 
308  kWh/m^  lower,  and  the  root-zone  temperature  during 
grov^rth  period  was  2-4  °C  lower  than  on  the  east-facing 
slope  at  2,185  m  above  sea  level  (a.s.l.)  (Turner  and  others 
1982).  The  mean  soil  temperature  at  10  cm  on  the  north- 
and  east-facing  slopes  from  June  to  September  was  3.5 
and  4.5  °C,  respectively  (Koike  and  others,  submitted). 

Plant  Materials 

Plants  studied  were  cembran  pine,  (provenance  2,050  m 
a.s.l.)  and  Norway  spruce  (provensmce  1,960  m  a.s.l.), 
raised  from  seeds  in  the  nursery  at  Birmensdorf  ZH  (550  m 
a.s.l.).  The  cembran  pines  were  transplanted  in  1975, 
and  the  Norway  spruces  in  1962,  to  the  experimental  site 
(2,185  m  a.s.l.).  Mean  tree  height  as  of  1988  on  the  north- 
facing  slope  and  east-facing  slope  was  40  cm  and  80  cm  in 
cembran  pine  and  36  cm  and  45  cm  in  Norway  spruce,  re- 
spectively. For  measurements  of  photos3mthesis,  attached 
shoots  vdthin  the  even-aged  needles  were  used.  Younger 
shoots  were  removed  when  the  photosynthesis  of  older 
shoots  was  measured.  The  cut  shoot  ends  were  sealed 
with  vaseline. 

Measurements 

The  gas  exchange  rate  was  measured  with  a  thermo- 
electrically  controlled  chamber  (Mini-cuvette  Walz,  FRG), 
which  was  coated  with  aluminium  foil  in  order  to  fully 
use  diffuse  light  (Koike  and  others,  submitted).  Photosyn- 
thesis was  determined  v^dth  two  infrared  gas  analyzers 
(Binos,  Leybold-Heraeus,  FRG).  Dew  point  mirrors  (Walz, 
FRG)  were  used  to  measure  the  absolute  air  hmnidity. 
Needle  temperatures  were  monitored  by  a  0.1  mm  chromel- 
constantan  thermocouple.  The  PPFD  above  the  needles 
was  measured  with  a  GaAsp  photodiode  (Hamamatsu 
Gil  18,  J)  after  calibration  vsrith  a  LiCor  quanttmi  sensor 
(USA).  A  COj  dispensing  apparatus  (Walz,  FRG)  pro- 
duced different  COg  concentrations  in  the  chamber. 


78 


Figure  1 — Cross  section  of  a  needle  of  a  cembran 
pine  (right)  and  a  NoPA'ay  spruce  (left). 


Anatomy  of  needles  was  observed  with  a  microscope 
after  needles  were  sectioned  with  a  microtome  (Leitz). 
Total  needle  surface  area  of  the  used  twig  was  estimated 
by  the  modified  glass  bead  method  (Davies  and  Benecke 
1980).  The  number  of  specimens  for  the  census  of  needle 
longevity  was  more  than  700  shoots  of  10  trees.  Dry 
weight  of  the  needles  was  determined  after  48  hours  dry- 
ing at  80  °C.  Nitrogen  concentration  in  the  needles  was 
analyzed  with  a  CHN  analyzer  (Rapid,  USA). 

RESULTS 

Needle  Characteristics 

The  anatomy  of  needles  of  cembran  pine  and  Norway 
spruce  was  similar  to  the  "homobaric"  leaves  (fig.  1). 
There  was  no  extension  of  vascular  bundle  sheaths  into 
the  mesophyll.  Needle  lifespans  of  cembran  pine  and 
Norway  spruce  were  9  and  12  years  for  the  north-facing 
slope  and  8  and  9  years  for  the  east-facing  slope,  respec- 
tively (fig.  2).  The  proportion  of  older  needles  of  both  spe- 
cies was  larger  on  the  north-facing  slope  than  on  the  east- 
facing  slope.  In  Norway  spruce,  the  frequency  of  younger 
needles  on  the  north-facing  slope  was  smaller  than  that 
on  the  east-facing  slope.  A  negative  correlation  between 
the  Ught-saturated  photosynthesis  and  needle  longevity 
was  found  in  both  species  (fig.  3). 


Physiological  Parameters 

There  was  a  positive  correlation  between  the  stomatal 
conductance  and  the  light-saturated  photosjoithesis  in 
both  species  (fig.  4).  No  difference  between  the  north-  and 
east-facing  slopes  was  foimd.  The  quantum  yield  and  car- 
boxylation  efficiency  (CE)  of  both  species  on  the  north- 
facing  slope  tended  to  be  lower  than  of  those  on  the  east- 
facing  slope  (table  1). 

The  maximimi  photo sjTithe sis  at  light  and  CO^  satura- 
tion (Pj^^)  of  both  species  on  the  north-facing  slope  was 
lower  than  that  on  the  east-facing  slope  (P  <  0.05).  The 
CE  of  cembran  pine  was  sHghtly  higher  than  that  of 
Norway  spruce.  The  nitrogen  concentration  in  needles 
of  cembran  pine  was  higher  than  in  needles  of  Norway 
spruce.  However,  there  was  no  statistical  difference  in 
both  species  between  the  slopes. 

DISCUSSION 

The  estimation  of    is  based  on  the  uniform  response 
of  the  stomata  of  a  leaf  (Terashima  1992).  The  anatomy 
of  needles  of  cembran  pine  and  Norway  spruce  was  simi- 
lar to  the  homobaric  leaf.  Non-uniform  stomatal  aperture 
in  needles  of  loblolly  pine  was  observed  only  when  needles 


Figure  2 — Frequency  of  needle  age  in  percent  of  Figure  3 — Photosynthesis  in  relation  to  needle  age 

total  needles  of  cembran  pines  (right)  and  Norway  in  cembran  pines  (left)  and  Norway  spruces  (right) 

spruces  (left)  growing  on  east-facing  and  north-  growing  on  east-facing  and  north-facing  slopes  at 

facing  slopes  at  timberline.  timberline. 


79 


8 


sz 

c 

w 
o 

o 
sz 


'E 

-54 

E 


1 

1 

•  North-facing  slope 

o 

uusi-Taciny  siop6 

1 

1 

0  40         80  0  40  80 

Stomotal  conductance (mmolrfi^s^) 

Figure  4 — Photosynthesis  in  relation  to  stomatal 
conductance  in  cembran  pines  (left)  and  Norway 
spruces  (right)  growing  on  east-facing  and  north- 
facing  slopes  at  timberline. 


were  treated  with  abscisic  acid,  not  with  the  low  tempera- 
ture in  roots  (Day  and  others  1991).  According  to  these 

facts,  we  can  estimate  the  CE  calculated  with  C.  values. 

1 

The  P^^  from  needles  of  the  north-facing  slope  was 
smaller  than  of  those  of  the  east-facing  slope,  while  needle 
lifespan  of  the  north-facing  slope  increased.  Many  physi- 
ological parameters  in  needles  from  the  north-facing  slope 
were  almost  the  same  as  those  on  the  east-facing  slope. 
However,  slightly  lower  P^^  in  both  species  from  the 
north-facing  slope  may  be  attributed  to  the  lower  soil 
temperature  (Day  and  others  1991;  Hasler  1982,  Turner 
and  others  1982).  The  lower  P     of  needles  from  the 

max 

north-facing  slope  may  be  compensated  for  by  the  longer 
lifespan  of  needles.  Tree  growth  on  the  north-facing  slope 
may  be  retarded  by  the  shortage  of  radiation  and  low  soil 
temperatures. 

ACKNOWLEDGMENTS 

We  thank  Dr.  H.  Turner  for  providing  the  climate 
data  of  the  Stillberg  site.  Prof.  W.  K.  Smith  for  collecting 
references,  and  Mrs.  M.  J.  Sieber  for  editing  the  English 
version  of  this  manuscript. 


REFERENCES 

Davies,  C.  E.;  Benecke,  U.  1980.  Fluidized  bed  coating 
of  conifer  needles  with  glass  beads  for  determination 
of  leaf  surface  area.  Forest  Science.  26(1):  29-32. 

Day,  T.  A.;  Heckathorn,  S.  A.;  DeLucia,  E.  H.  1991.  Limi- 
tations of  photosynthesis  in  Pinus  taeda  L.  (loblolly 
pine)  at  low  soil  temperatures.  Plant  Physiology.  96: 
1246-1254. 

Hasler,  R.  1982.  Net  photosynthesis  and  transpiration 
of  Pinus  montana  on  east  and  north  facing  slopes  at 
alpine  timberline.  Oecologia.  54:  14-22. 

Koike,  T.;  Hasler,  R.;  Matyssek,  R.;  Item,  H.  Seasonal 
changes  in  the  photosynthetic  capacity  of  Larix  decidua 
and  Pinus  cembra  planted  on  contrasting  slopes  at  the 
timberline  at  Stillberg  (Davos),  eastern  Switzerland. 
Trees,  submitted. 

Kuoch,  R.;  Amiet,  R.  1970.  Die  Verjiingung  im  Bereich  der 
oberen  Waldgrenze  der  Alpen.  Eidgenossische  Anstalt 
fiir  das  forstliche  Versuchswesen,  Mitteilungen.  46: 
159-328. 

Livingston,  N.  J.;  Black,  T.  A.  1987.  Stomatal  characteris- 
tics and  transpiration  of  three  species  of  conifer  seed- 
lings planted  on  a  high  elevation  south-facing  clear-cut. 
Canadian  Journal  of  Forest  Research.  17:  1273-1282. 

Schonenberger,  W.;  Frey,  W.  1988.  Untersuchungen  ziar 
Okologie  imd  Technik  der  Hochlagenaufiforstimg. 
Forschungsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fur  Forstwesen. 
139:  735-820. 

Terashima,  L  1992.  Anatomy  of  non-uniform  leaf  photo- 
synthesis. Photos)aithesis  Research.  31(3):  195-212. 

Turner,  H.;  Hasler,  R.;  Schonenberger,  W.  1982.  Contrast- 
ing microenvironments  and  their  effects  on  carbon  up- 
take and  allocation  by  yoimg  conifers  near  alpine  tree- 
line  in  Switzerland.  In:  Waring,  R.  H.:  Carbon  uptake 
and  allocation  in  subalpine  ecosystems  as  a  key  to  man- 
agement: Proceedings  of  the  lUFRO  Workshop, 
Pl.07-00,  Corvalhs,  OR:  Oregon  State  University: 
22-30. 


Table  1 — Physiological  parameters  of  cembran  pine  and  Norway  spruce  growing  on  an  east-  and  a  north-facing  slope  of  an 
avalanche  gully  at  Stillberg  (2,185  m  a.s.l.) 


Parameter 


Cembran  pine 


E  slope 


N  slope 


Norway  spruce 


E  slope 


N  slope 


Quantum  yield  0.014  ±0.004 
(mol  COj'mol/quanta) 

Carboxylation  efficiency  0.024  ±  0.009 
(^mol•m-2•^ba^^' 


(mol»m-2»s-^) 


8.30  ±  1.51 


Nitrogen  concentration 
(mg»g-^)  12.26  ±  0.90 


0.009  ±  0.002 

0.022  ±  0.009 

5.38  ±  0.77 

13.27  ±  0.91 


ns 


ns 


(1) 


ns 


0.012  ±  0.006       0.006  ±  0.002 


0.017  ±  0.002       0.018  ±  0.008 


6.87  ±  1.61 


10.43  ±  1.87 


4.93  ±0.12 


9.26  ±  1.22 


ns 


ns 


(1) 


ns 


'P<0.05. 


80 


THE  BROAD-LEAVED  KOREAN  PINE 
FOREST  IN  CHINA 

Luo  Ju  Chun 


Abstract — Korean  pine  (Pinus  koraiensis)  is  a  tree  species  of 
high  economic  value.  Its  natural  distribution  area,  the  character- 
istics of  distribution  in  the  broad-leaved  Korean  pine  forest  re- 
gion of  China,  the  biological  and  ecological  characteristics  of  Ko- 
rean pine,  the  forest  types  of  Korean  pine,  and  the  succession 
pattern  of  the  Korean  pine  forest  communities  have  been  studied. 
The  results  are  significant  to  conservation  and  sustaining  use  of 
Korean  pine. 


Korean  pine  (Pinus  koraiensis)  is  well  known  for  its 
high-quality  timber  that  can  be  used  for  many  different 
purposes.  Undoubtedly,  Korean  pine  is  one  of  the  most 
important  tree  species  in  Chinese  forests. 

Except  for  a  few  pure  Korean  pine  forests,  the  species 
often  occurs  in  the  "conifer-broad-leaved  tree  mixed-forest 
communities"  vnth  many  species  of  deciduous  broad- 
leaved  trees  and  other  conifers,  which  is  the  most  repre- 
sentative vegetation  formation  in  the  northeastern  humid 
zone  in  China. 

NATURAL  DISTRIBUTION  AREA 

Korean  pine — a  relict  species  from  the  Tertiary — 
mainly  occurs  in  northeast  Asia  and  in  a  narrow  zone 
along  the  west  coast  of  the  Japan  Sea.  This  zone  includes 
the  southern  part  of  eastern  Russia,  the  eastern  part  of 
northeast  China,  North  Korea,  and  the  center  of  Honshu 
in  Japan.  The  main  natural  range  of  Korean  pine,  how- 
ever, is  in  China.  Korean  pine  is  a  representative  species 
in  Dahurian  flora,  and  also  is  the  main  forest  tree  species 
in  the  eastern  momitains  of  northeast  China. 

In  China,  the  broad-leaved  Korean  pine  forest  is  found 
from  40°  15'  N.  to  50°20'  N.  and  from  126°  E.  to  135°  E., 
east  of  Song-nen  plain  and  north  of  Song-liao  plain.  In 
the  south  it  reaches  Dan  dong,  in  the  north  Hei-Ho 
coimty.  The  whole  sirea  takes  a  crescent  shape  covering 
Lesser  Xingan  Ling,  Wan  Da  Mountain,  Zhang  Guang  Cai 
Ling,  Lao  Ye  Ling,  and  Chang  Bai  Mountain.  Most  of 
these  mountains  are  trending  from  northeast  to  south- 
west. They  are  characterized  by  gentle  topography  and 
low  elevation,  not  higher  than  1,300  m.  There  are  many 
different  tree  species,  including  many  quite  valuable 
broad-leaved  trees.  Their  exploitation  began  more  than 
100  years  ago.  Their  timber  output  still  makes  up  one- 
fourth  of  total  timber  production  in  China. 

The  forest  region,  which  is  influenced  by  the  Japan  Sea, 
is  characterized  by  temperate-zone  monsoon  climate.  The 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Luo  Ju  Chun  is  Professor  of  Ecology  and  Forest  Management,  Forestry 
Resource  College,  Forestry  University,  Beijing,  China. 


warm  and  moist  climate  is  favorable  to  growth  of  Korean 
pine.  Annual  average  temperature  is  0-6  °C;  the  growing 
season  is  about  4-5  months.  Temperatvu"e  svun  (>10  °C) 
is  about  2,500-3,200  °C.  At  higher  latitude,  temperature 
is  low,  vyinter  is  long,  and  simimer  short.  Winter  covers 
5  months.  January  is  the  coldest  month;  the  absolute 
minimum  temperature  is  -40  to  -30  °C.  In  July,  which  is 
the  warmest  month,  mean  temperature  reaches  20-24  °C. 
Winter  is  characterized  by  thick  snow  cover  and  mean 
temperatures  ranging  from  -28  to  -14  °C.  Annual  pre- 
cipitation amoimts  to  600-1,100  mm  £ind  decreases  from 
south  to  north;  maximum  is  in  summer  (June  to  August). 
The  annual  relative  hiunidity  is  about  60-70  percent. 
Hiunidity  and  temperature  are  favorable  to  growth  of 
Korean  pine  and  other  trees. 

The  soil  in  Korean  pine  forests  is  dark  brown  forest  soil 
vfith  a  thick  hvunus  layer,  and  it  is  very  fertile. 

CHARACTERISTICS  OF 
VEGETATION  DISTRIBUTION 

Horizontal  Distribution 

Because  of  the  wide  range  of  Korean  pine  forest,  the  cli- 
mate is  quite  different  from  north  to  south  in  the  distribu- 
tion area.  Consequently,  some  differences  in  flora  and 
plant  commvuiities  become  apparent.  The  distribution 
area  can  be  divided  into  two  subzones:  northern  and 
southern  temperate.  The  demarcation  is  from  Don-g  Nin 
in  the  eastern  part  of  Jing  Bo  Lake  to  Ji  Lin  district  until 
its  west  boiindary;  it  seems  to  nm  along  the  parallel  of 
44  °N.  The  differences  in  natiiral  conditions  and  vegeta- 
tion can  be  described  as  follows: 

1.  In  the  northern  subzone,  climate  is  cold,  the  growing 
season  is  only  about  100-120  days,  and  annual  rainfall 
£imounts  to  500-700  mm;  the  southern  subzone  is  charac- 
terized by  a  warm  and  hvunid  climate  and  a  growing  sea- 
son of  130-150  days.  Annual  precipitation  is  about  700- 
1,100  mm. 

2.  In  the  northern  subzone,  more  conifers  typical  of  the 
cold-temperate  zone  (for  example,  Picea  koraiensis,  P. 
jezoensis,  and  Abies  nephrolepis)  occur  in  Korean  pine 
forests.  Pinus  pumila  can  be  found  in  the  subalpine.  In 
addition,  Quercus  mongolica  is  more  common  on  expo- 
siu-es  to  Bun.  Larix  gmelinii  forests,  mixed  with  Betula 
platyphylla,  occur  in  the  lowlands.  Vaccinium  vitis- 
idaea,  Ledum  palustre,  Vaccinium  uliginosum,  Betula 
fructicosa — representatives  of  Dahurican  flora — are  typi- 
cal of  larch  forest  undergrowth.  In  the  southern  subzone, 
there  are  more  representatives  of  Chang  Bai  flora,  such  as 
Picea  jezoensis  var.  komarovii,  Abies  holophylla,  and 
Larix  olgensis. 


81 


3.  In  the  southern  subzone  there  are  more  plant  species 
than  in  the  north.  The  forest  often  consists  of  30-40  woody 
species  forming  three  layers.  In  the  southern  subzone 
there  are  twice  as  many  broad-leaved  tree  species  in 
Korean  pine  forests  than  in  the  northern  subzone;  the 
genus  Acer,  for  instance,  is  represented  by  only  five  spe- 
cies in  the  north,  compared  to  12  species  in  the  south. 
The  percentage  of  valuable  broad-leaved  trees  (such  as 
Franus  mandshurica,  Juglans  mandshurica,  Phello- 
dendron  amurense,  Tilia  amurensis,  Ulmus  propingua) 
in  Korean  pine  forests  is  higher  in  the  southern  than  in 
the  northern  subzone.  These  broad-leaved  trees  usually 
grow  together  with  Betula  costata  and  Acer  mono  and  rep- 
resent more  than  50  percent  of  the  species  of  southern 
Korean  pine  forests.  Moreover,  Carpinus  cordata  and 
Fraxinus  rhynchophylla  are  very  commom  in  the  south, 
but  occur  rarely  in  the  north.  In  addition,  there  are  more 
than  10  vine  plant  species  in  the  southern  subzone,  but 
only  five  or  six  in  the  north. 


1400  -r 


0-^ 


Figure  1 — Vertical  distribution  pattern  of  veg- 
etation in  Lesser  Xin  An  Ling  (southern  expo- 
sures): 1a,  River  head-valley-conifer  forest  with 
larch,  spruce,  fir;  1b,  Bank  depression,  broad- 
leaved  forest  with  Ulmus  propinqua,  Fraxinus 
mandshurica,  Pliellodendron  amurense;  2,  Oak 
forest  {Quercus  mongolica)  mixed  with  Betula 
dahurica,  Tilia  mandshurica;  3,  Broad-leaved 
Korean  pine  forest  mixed  with  spruce,  fir, 
Betula  costata,  Tilia  amurensis,  Ulmus 
laciniata;  4,  Spruce-fir  forest  belt;  5,  Subalpine 
Betula  ermanii-Pinus  pumila  elfin  forest  belt. 


Vertical  Distribution 

The  vertical  distribution  pattern  of  vegetation  in  Lesser 
Xin  An  Ling  (fig.  1)  is  a  typical  example  of  the  northern 
subzone: 

1.  River  valley  forest:  It  is  composed  of  Picea  jezoensis, 
Abies  nephrolepis,  Larix  gmelinii,  Juglans  mandshurica, 
Ulmus  propinqua,  Populus  ussuriensis,  and  other  species. 

2.  Oak  {Quercus  mongolica)  broad-leaved  forest  belt 
(150-300  m):  It  is  secondary  forest  that  established  after 
destruction  of  virgin  conifer-broad-leaved  mixed  forest, 
and  it  occurs  usually  on  shallow  soils  on  s\in-exposed 
slopes. 

3.  Broad-leaved  Korean  pine  forest  belt  (250-650  m): 
Korean  pine  is  the  dominant  species;  spruces  and  firs  oc- 
casionally occur;  the  broad-leaved  trees  make  up  less  than 
30  percent  of  the  species.  The  volume  of  standing  timber 
is  the  highest  (300-600  m^/ha)  in  the  northern  forest  re- 
gion. Korean  pine  grows  on  exposures  to  south  and  south- 
west, on  mountain  ridges,  and  on  hill  tops. 

4.  Spruce-fir  forest  belt  (650-1,000  m):  Picea  korai- 
ensis,  Picea  jezoensis,  and  Abies  nephrolepis  are  the  main 
tree  species. 

5.  Subalpine  moss-elfin  forest  belt  (1,000-1,080  m): 
The  climate  is  cold  and  humid;  wind  is  strong.  The  elfin 
forest  is  composed  of  Betula  ermanii.  This  species  exhib- 
its dwarf  growth  forms  not  taller  than  4-6  m.  Pinus 
pumila  grows  in  the  forest  understory.  The  forest  floor 
is  covered  mainly  by  moss. 

The  vertical  distribution  pattern  of  vegetation  in  the 
southern  subzone  can  be  illustrated  by  the  situation  of 
Chang  Bai  Mountain  (fig.  2): 

1.  Broad-leaved  forest  belt  (250-500  m  a.s.l.):  It  is 
formed  by  Quercus  mongolica,  Betula  platyphylla,  Betula 
dahurica,  and  Populus  davidiana.  It  is  a  typical  secon- 
dary forest  following  destruction  of  broad-leaved  Korean 
pine  forest.  Locally,  basswood,  maple,  elm,  walnut,  and 
corktree  form  mixed  forests.  At  lower  elevations,  these 
broad-leaved  forests  are  scattered. 

2.  Broad-leaved  Korean  pine  forest  belt  (500-1,200  m): 
Korean  pine  is  the  main  species,  and  is  mixed  with  vari- 
ous broad-leaved  trees:  Betula  costata,  Carpinus  costata, 
Ulmus  laciniata,  Acer  mono,  Tilia  amurensis,  Quercus 
mongolica,  especially  the  valuable  broad-leaved  tree  spe- 
cies Ulmus  propinqua,  Phellodendron  amurense,  Juglans 
mandshurica,  Fraxinus  mandshurica,  and,  at  valley  bot- 
toms, some  spruce  and  fir  can  be  found. 

3.  Spruce-fir  forest  belt  (1,200-1,800  m):  It  is  formed 
by  Picea  jezoensis,  Picea  koraiensis,  and  Abies  nephrolepis, 
with  additional  Betula  ermanii  and  Larix  olgensis  in  the 
upper  part  of  the  belt,  or  with  Korean  pine,  Betula 
costata,  Ulmus  laciniata,  and  Tilia  amurensis  in  its  lower 
part. 

4.  Subalpine  Betula  ermanii  mossy  elfin  forest  belt  and 
subalpine  meadow  belt  (1,800-2,100  m):  These  belts  are 
characterized  by  cold  and  windy  climate,  steep  topogra- 
phy, and  shallow  soils.  Betula  ermanii  forests  alternate 
with  vast  meadow  areas. 

5.  Alpine  tundra  belt  (>2,100  m):  Vegetation  consists 
of  different  dwarf  shrub  species,  mosses,  and  lichens. 


82 


3000  -r 


2500 


2000  - 


(0 

5 1600  - 
0) 


1000  - 


500  - 


0-^ 


Figure  2 — Vertical  zonal  spectrum  of  vegetation 
in  Chang  Bai  Mountain:  1 ,  Secondary  deciduous 
broad-leaved  forest;  2,  Conifer-broad-leaved 
mixed  forest  (broad-leaved  Korean  pine  forest); 
3,  Coniferous  forest  belt:  3a,  Spruce-fir-Korean 
pine  forest  subbelt;  3b.  Spruce-fir  coniferous  for- 
est subbelt;  4,  Subalpine  Betula  ermanii-e\i\n  for- 
est belt,  and  subalpine  meadow  belt;  5,  Alpine 
tundra  belt. 


As  far  as  we  know,  the  higher  the  latitude  the  lower  is 
the  upper  limit  of  Korean  pine.  In  the  southern  subzone 
Korean  pine  occurs  up  to  1,200  m  a.s.l.;  in  the  north  its 
uppermost  altitudinal  limit  is  located  at  about  650  m. 

BIOLOGICAL  AND  ECOLOGICAL 
CHARACTERISTICS 

Korean  pine  is  a  long-lived  species.  Its  lifespan  is  300- 
400  years,  sometimes  500  years.  Tree  height  usually 
reaches  30  m.  The  diameter  can  reach  1  m.  Korean  pine 
is  one  of  the  valuable  sawtimber  trees  in  the  world.  It 
grows  slowly  in  the  first  5-8  years.  After  the  age  of  10  its 
growth  rate  increases  rapidly.  Its  diameter  growth  begins 
to  accelerate  at  the  age  of  12.  The  height  growth  becomes 
more  rapid  after  16-18  years.  Accelerated  growth  contin- 
ues vmtil  the  age  of  30.  Korean  pine  is  a  tree  species 
growing  fastest  at  middle  age.  In  general,  growth  rate  of 
Korean  pine  increases  earlier  in  plantations  than  in  natu- 
ral forests.  In  clear  cuttings,  the  radial  growth  of  8-year- 
old  trees  is  2.6  times  higher  than  in  nattu-al  forests,  and 
height  growth  is  twice  as  much  as  in  trees  at  the  same 
age  that  became  naturally  established. 


Secondary  growth  of  Korean  pine  is  common  in  autiunn, 
but  it  appears  mainly  in  trees  less  than  10  years  old. 
After  80  years  growth,  Korean  pine  reaches  a  height  of 
18-20  m.  At  this  age  the  tree  top  often  begins  to  fork  and 
bear  cones.  Obviously,  the  forking  is  closely  related  to 
cone  production. 

Korean  pine  has  shallow  roots.  Its  taproot  is  imderde- 
veloped  and  degenerates  easily.  Thus,  seed  trees  left  on 
clear  cuttings  are  sensitive  to  windbreak. 

Most  Korean  pine  trees  begin  to  produce  cones  at  the 
age  of  80-100  years.  Nevertheless,  cones  can  also  be 
found  on  30-year-old  trees.  In  plantations,  however, 
even  20-year-old  trees  are  able  to  bear  cones. 

There  are  two  ecot5T3es  of  Korean  pine:  one  is  the 
leptodermis  form  displaying  thin  bark  with  small  and 
shallow  scale  or  long  strip  lobes.  It  forks  less,  grows 
faster  in  height,  and  produces  better  timber.  The  other 
is  the  pachidermis  form  characterized  by  deep  and  large, 
long,  square-shape  lobes  on  the  tree  trunk.  It  forks  more, 
grows  slower  in  height,  and  the  timber  quality  is  not  as 
good  as  in  the  leptodermis  type.  The  leptodermis  type 
should  be  used  for  afforestation. 

Korean  pine  belongs  to  thermophile  species;  its  vegeta- 
tive growth  requires  6-7  °C  of  the  lowest  average  tempera- 
tiu-e  and  the  most  suitable  temperature  is  14-16  °C  in  the 
yoimg  stage.  The  relative  moistvu^e  it  demands  is  70  per- 
cent and,  during  the  period  of  most  rapid  growing,  rela- 
tive hvunidity  of  about  70  percent  is  required.  Korean 
pine  is  intolerant  to  high  temperature  during  vegetative 
growth.  If  average  temperature  exceeds  15-16  °C,  height 
growth  will  decrease  or  even  stop.  On  the  other  hand,  it 
will  not  be  affected  by  winter  temperatiires  as  low  as 
-50  °C. 

At  seedling  stage  a  little  shading  is  favorable  to  Korean 
pine,  althovigh  it  can  tolerate  full  siinlight.  Its  light  re- 
quirement increases  with  age.  While  3  years  old,  seed- 
lings require  60-70  percent  of  incoming  solar  radiation. 
After  the  fifth  year,  young  trees  become  totally  shade 
intolerant. 

Korean  pine  grows  best  on  thick,  moist,  fertile,  and  well- 
drained  soils  with  pH  5-6,  but  it  can  also  exist  on  shallow 
dry  soil,  poor  in  nutrients.  It  grows  badly  on  soils  rich  in 
clay.  However,  Korean  pine  requires  appropriate  soil  mois- 
ture. In  bogs  and  on  sites  with  stagnant  water,  growth  is 
hampered,  and  Korean  pine  is  rarely  seen  there. 

PRINCIPAL  FOREST  TYPES 

Mu  Dan  Jiang  forest  region,  the  distributive  center  of 
Korean  pine,  may  serve  as  an  example  to  classify  Korean 
pine  forest  into  the  following  forest  types: 

Steep-Slope  Carex  calltnichos-Korean  Pine 
Forest — This  type  occurs  on  simny  exposures  or  narrow 
ridges  at  an  altitude  of  800-1,000  m.  These  usually  are 
poor  sites.  The  conmiunity  can  be  divided  into  two  layers. 
The  forest  canopy  is  formed  by  Korean  pine  (90  percent) 
and  Quercus  mongolica  or  Tilia  amurensis  (10  percent), 
which  have  a  higher  crown  density  than  Korean  pine.  In 
the  second  layer  there  is  some  Korean  pine,  spruce,  fir, 
and  linden.  The  growing  stock  is  400-500  m^/ha.  The 
dominant  species  on  the  forest  floor  are  Carex  calltnichos 
and  C.  ussuriensis. 


83 


CoryluS'Carex-Korean  Pine  Forest— This  type  is 
found  on  mountains  with  gentle  slopes  at  altitudes  of 
700-900  m.  As  to  favorability  to  tree  growth,  these  sites 
can  be  considered  intermediate.  Korean  pine  prevails 
(60-70  percent)  followed  by  Tilia  amurensis  (20  percent), 
and  some  Abies  nephrolepis,  Picea  jezoensis,  Ulmus 
propinqua,  Acer  mono,  and  Betula  costata.  The  growing 
stock  is  about  400-500  m^/ha.  There  are  more  tree  species 
associated,  such  as  Acer  sp.  There  is  more  undergrowth, 
and  the  main  species  is  Corylus  mandshurica.  On  the  for- 
est floor  about  30  species  are  to  be  found.  Carex  sidero- 
sticta  and  ferns  are  the  most  common  plants. 

Fern-Spruce-Fir-Korean  Pine  Forest — This  type 
occurs  on  the  lower  gentle  slopes  (inclination  about  10°) 
of  the  mountains  or  along  side  streams  at  an  altitude  of 
500-800  m.  The  community  construction  is  complex.  The 
canopy  layer  is  formed  by  Korean  pine  and  some  Populus 
ussuriensis  and  Picea  jezoensis.  The  second  layer  is  com- 
posed of  Picea  jezoensis  and  P.  koraiensis  (50  percent), 
Abies  nephrolepis  (20  percent),  and  Tilia  amurensis 
(20  percent).  Other  species  are  Betula  costata,  Ulmus 
laciniata,  Acer  mono,  Phellodendron  amurense,  Juglans 
mandshurica,  Fraxinus  mandshurica,  and  many  kinds  of 
maple.  Because  of  better  site  conditions,  growing  stock  is 
higher  (250-350  m^/ha  for  the  first  story  and  200  m^/ha  for 
the  second  story).  The  undergrowth  is  abundant  with 
over  20  species;  about  30  species  occur  on  the  forest  floor. 
Ferns  are  dominant  (such  as  Dryopteris). 

Fern-Moss  Korean  Pine  Forest — This  type  is  found 
in  the  lower  part  of  mountains  or  in  broad  valleys;  slope 
inclination  is  0-5°.  Two  stories  can  clearly  be  distin- 
guished: Pinus  koraiensis  (80  percent)  and  a  few  Picea 
jezoensis,  Fraxinus  mandshurica,  Ulmus  propinqua,  Acer 
mono,  and  Tilia  amurensis  form  the  forest  canopy.  In  the 
second  story  there  is  more  Picea  jezoensis,  Betula  costata, 
Acer  mono,  Abies  nephrolepis,  Tilia  amurensis,  and  a  few 
Juglans  mandshurica,  Fraxinus  mandshurica,  Ulmus 
propinqua,  Betula  platyphylla,  and  many  kinds  of  maple. 


So  the  tree  species  are  very  abundant.  This  forest  type 
is  typical  of  the  sites  most  favorable  to  tree  growth.  The 
growing  stock  can  reach  600  m^/ha.  The  imdergrowth 
mainly  consists  of  Lonicera  sp.,  Viburnum  sp.,  and  Sor- 
baria,  and  is  well  developed.  Carex  and  multiple  fern  and 
luxuriant  moss  cover  are  common  throughout  the  area. 

COMMUNITY  SUCCESSION 
PATTERNS 

Two  succession  patterns  for  the  broad-leaved  Korean 
pine  communities  are  obvious: 

Succession  Without  Disturbance — Natural  regen- 
eration is  poor  under  virgin  Korean  pine  forest  cover,  al- 
though a  comparatively  large  quantity  of  seedlings  may 
appear  after  a  seed  year.  However,  they  will  only  be  able 
to  grow  up  if  the  canopy  is  cut  thin.  Because  Korean  pine 
lives  longer  than  other  coniferous  and  broad-leaved  trees, 
and  tolerates  shade  for  longer  period,  it  tends  to  form 
multigeneration  and  multistoried  forests. 

Succession  After  Disturbance — ^After  destruction  of 
broad-leaved  Korean  pine  forest  the  succession  goes  on  as 
shown  by  the  following  sequences: 

Selective  cutting  -»  conifer-broad-leaved  mixed  forest 
in  which  main  elements  are  broad-leaved  trees.  Protec- 
tion    broadleaved-Korean  pine  forest. 

Clear  cutting  or  fire  ->  secondsiry  bare  land.  Protec- 
tion    grass  and  shrub  communities  conifer-hardwood 
mixed  forests  in  which  main  elements  are  hardwood  trees 
->  conifer-broad-leaved  mixed  forest  in  which  main  ele- 
ments are  coniferous  trees     broad-leaved-Korean  pine 
forest. 

To  promote  the  reestablishment  of  Korean  pine  forest, 
management  should  be  adjusted  to  this  succession 
process. 


84 


CLIMATES  WHERE  STONE  PINES 
GROW,  A  COMPARISON 

Tad  Weaver 


Abstract— While  stone  pine  climates  are  similar,  species 
adapted  to  relatively  moderate  climates  may  be  excluded  from 
the  ranges  of  congeners  by  more  severe  climates,  and  species 
with  longer  warm-moist  growing  seasons  are  probably  more  pro- 
ductive than  congeners.  Absolute  low/summer  average/absolute 
high  temperatures  for  stone  pines  listed  in  order  of  increasing 
absolute  low  temperature  are  Pinus  sibirica  (-55/13/37  °C), 
P.pumila  (-52/9/36  °C),  P.  koraiensis  (-42/11/36  °C),  P.  albi- 
caulis  (-34/9/29  °C),  and  P.  cembra  (-23/8/27  °C).  The  Walter 
drought  index  shows  little  stress  in  stone  pine  forests  despite 
large  differences  in  summer/winter  precipitation:  in  order  of 
increasing  summer  rainfall,  precipitation  is  P.  albicaulis  (102/ 
829mm),  P.  pumiVa  (142/264mm),  P.  sibirica  {187 /245mm), 
P.  cembra  (323/616mm),  P.  koraiensis  (394/242mm).  Estimated 
thawed-soil  growing  season  increases  from  P.  albicaulis  (4.5mo), 
through.  P.  pumila  (4.6mo),  P.  sibirica  (5.5mo),  and  P.  cembra 
(6.3mo)  to  P.  koraiensis  (7.8mo);  growing  seasons  of  the  first 
three  trees  could  be  shortened  by  drought. 


Stone  pines  grow  in  most  of  the  boreal  zone  (fig.l) 
(Fullard  and  Darby  1964;  Lanner  1990;  Mirov  1967). 
Pinus  pumila  and  P.  sibirica  occupy  vast  far-northern 
(50  to  70°  N.  latitude)  areas  in  northeastern  Asia;  P.  sibir- 
ica ranges  from  just  west  of  the  Urals  (55°  E.  longitude) 
halfway  to  the  Pacific  (115°  E.  longitude)  and  P.  pumila 
ranges  from  115°  E.  longitude  to  the  Pacific  (165°  E.  longi- 
tude). P.  koraiensis  grows  at  the  boreal-deciduous  forest 
transition  in  eastern  China  (120  to  150°  E.  longitude  and 
45  to  55°  N.  latitude  with  outliers  to  35°  N.  latitude  at 
high  altitude).  Paralleling  P.  koraiensis,  P.  albicaulis 
occupies  the  Rocky-Cascade-Sierra  Mountain  chains  of 
North  America  from  55°  N.  latitude  southward  to  45°  N. 
latitude  with  outhers  to  near  35°  N.  latitude.  Pinus 
cemba  appears  in  the  Alps  at  approximately  45°  N.  lati- 
tude. Glaciated  parts  of  the  conifer  forest  zone— both  in 
northern  Europe  and  North  America— lack  stone  pines 
(fig.l). 

Due  to  their  common  ancestry  and  common  boreal 
forest  habitat,  one  expects  the  trees  to  occupy  similar  cli- 
mates. This  paper  tests  that  hypothesis  by  comparing  cli- 
mates occupied  by  the  trees  with  respect  to  20factors  that 
may  be  important  in  determining  the  ranges  and  produc- 
ivities  of  the  species.  The  information  may  be  especially 
useful  for  predicting  the  success  of  introductions— from 
one  region  to  another— of  stone  pines,  their  tree  associates, 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Statxis  of  Our  Knowledge,  St.Moritz, 
Switzerland,  September5-ll,  1992. 

Tad  Weaver  is  Plant  Ecologist,  Biology  Department,  Montana  State 
University,  Bozeman,  MT  59717. 


and  to  some  degree,  even  their  herb,  crj^jtogam,  and  ani- 
mal associates.  Such  introductions  might  be  of  interest 
as  either  producers  or  pests. 

METHODS 

Climates  of  environments  dominated  by  closely  related 
stone  pines  (Lanner  1990;  Mirov  1967)  were  compared. 
The  comparisons  are  nonstatistical,  since  the  objective 
was  to  discuss  biological  responses  rather  than  the  cli- 
mates themselves. 

The  climates  were  characterized  by  choosing  three  to 
four  stone  pine  stands  at  which  temperature  and  precipi- 
tation were  regularly  measured  (a  complete  sample  of 
accessible  data);  summarizing  the  data  for  periods  of  10 
years  (P.  albicaulis  and  P.  cembra)  or  longer  (unspecified); 
and  calculating  means  for  selected  parameters.  The  sta- 
tions studied  are  Usted  here— from  west  to  east— each 
with  its  approximate  latitude,  longitude,  and  altitude  (m). 
Note  that  P.  albicaulis  and  P.  cembra  grow  significantly 
higher  than,  and  to  the  south  of,  the  Asian  pines.  Aster- 
isked Siberian  stations  lack  most  temperature  data.  Sci- 
entists who  helped  identify  weather  stations  in  stone  pine 
climates  are  listed  in  the  acknowledgments. 

Pinus  albicaulis  stations  (USDC  1961-80)  were  Crater 
Lake,  OR  (43°  N.,  122°  W.,  1,990m),  Ellery  Lake,  CA 
(38°  N.,  119°  W.,  2,940m),  Old  Glory,  BC  (49  °  N.,  119°  W., 
1,008m),  and  Kings  Hill,  MT  (46  °  N.,  110°  W.,  2,225m). 
Pinus  cembra  stations  (contributed  by  W.Tranquillini) 
were  Haggen  (47°  N.,  11°  E.,  1,800m),  Obergurgl  (47  °  N., 
11°  E.,  2,070m),  and  Patscherkofel  (48  °  N.,  11°  E., 
1,952m).   Pinus  sibirica  stations  (Muller  1982)  were 
Serov  (60°  N.,  61°  E.,  132m*),  Surgut  (61  °  N.,  73°  E., 
40m),  Kolpasevo  (58  °  N.,  83°  E.,  76m*),  Jenisejsk 
(58°  N.,  92°  E.,  78m*),  Krasnojarsk  (56  °  N.,  93°  E., 
15m),  Tura  (64  °  N.,  100°  E.,  130m*),  Irkutsk  (52  °  N., 
104°  E.,  468m),  and  Kirensk  (58  °  N.,  108°  E.,  256m). 
Pinus  pumila  stations  (Muller  1982)  were  Vitujsk  (64°  N., 
122°  E.,  107m*),  Jakutsk  (62  °  N.,  130°  E.,  100m*), 
Verchojansk  (68°  N.,  133°  E.,  137  m),  Ochotsk  (59°  N., 
143°  E.,  6  m),  Zyranka  (65°  N.,  151°  E.,  43m*), 
Petropavlovsk-Kamcatskij  (53°  N.,  159°  E.,  32m),  Apuka 
(65°  N.,  170°  W.,  10m*),  and  Anadyr  (65  °  N.,  178°  E., 
62m).  Pinus  koraiensis  stations  (contributed  by  Luo 
Ju  Chun  and  Zhao  Shidong)  were  AnTu  (43°  N.,  128°  E., 
591m),  Chargbei-ShenYang  (42  °  N.,  128°  E.,  738m)  and 
YiChun  (48°  N.,  129°  E.,  231m). 

Parameters  studied  describe  climates  with  respect  to 
temperature,  precipitation,  and  growing  season.  The  fol- 
lowing three  paragraphs  list  these  parameters  and  outline 
the  rationales  for  their  choice;  the  rationales  are  devel- 
oped further  in  Weaver  (1993)  and  in  the  following 
discussion. 


85 


B 

7^ — 

•1 

Figure  1 — Distribution  of  stone 
pines  in  relation  to  space,  conifer 
forests,  and  pleistocene  glaciation. 
North  America  and  Eurasia  are 
seen  from  a  polar  view.  Shading 
locates  the  conifer  zone  on  the  two 
continents  (Fullard  and  Darby 
1964).  Areas  within  the  dotted  lines 
received  pleistocene  glaciation 
(Denton  and  Hughes  1981). 
Ranges  of  the  stone  pines  (Mirov 
1967)  are  outlined  by  solid  or 
dashed  lines:  A  =  Pinus  cembra, 
B  =  P.  sibirica,  C  =  P.  pumila, 
D  =  P.  koraiensis,  E  =  P.  albicaulis. 
The  absence  of  stone  pines  in  the 
conifer  zone  of  northwestern 
Eurasia  and  parts  of  North  America 
could  be  due  to  glaciation. 


Ten  temperature  parameters  were  compared.  Midwin- 
ter frost  danger  was  described  by  the  long-term  minimum 
temperature  (the  absolute  low),  the  average  January 
minimum,  and  the  average  January  maximum.  Extreme 
fall  and  spring  frosts  are  represented  by  the  absolute  low 
in  the  first  and  last  months  of  winter  (that  is,  the  absolute 
lows  in  the  months  with  OC  average  air  temperature; 
Weaver  1994).  The  average  growing-season  temperature 
was  calculated  across  those  months  when  average  air 
temperatures  were  above  0  °C.  This  average  was  recalcu- 
lated using  "temperature  growth  support  units  =  Q"  to 
account  for  the  rise  in  rates  of  metabolic  processes  with 
rising  temperature;  these  points  lie  on  a  curve  defined 
by  0  °C  =  0,  1  °C  =  1,  11  °C  =  2,  21  °C  =  3,  31  °C  =  4,  and 
41  °C  =  5  (Weaver  1994).  Summer  highs  were  represented 
by  the  average  July  minimum,  average  July  maximum, 
and  long-term  high. 

Seven  precipitation  parameters  were  compared. 
October- June  precipitation  was  measured  to  determine 
whether  soil  was  moist  in  winter  and  at  the  opening  of 


the  growing  season.  Summer  precipitation  was  indexed 
by  July-September  precipitation  and  by  precipitation  in 
the  wettest  and  driest  month  in  that  period.  Since  plants 
require  water,  not  precipitation,  months  with  a  positive 
water  balance  were  estimated  with  the  Walter  index 
(Nielson  1986;  Walter  1973),  which  assumes  that  for  ev- 
ery 2  °C  rise  in  average  temperature  1  mm  of  precipita- 
tion will  be  evapotranspired.  Two  derivatives  were  stud- 
ied: drought  months  (duration  of  periods  with  a  negative 
water  balance)  and  drought  magnitude  (total  estimated 
deficit).  While  the  Walter  index  tends  to  underestimate 
drought  (Stephenson  1990),  it  may  be  adequate  in  these 
cool  climates.  Poikilohydric  (Larcher  1975)  organisms 
may  be  more  dependent  on  a  number  of  growing  hours  set 
by  the  number  of  rain  days  than  on  total  precipitation; 
rain  days  per  month  were  therefore  recorded. 

Four  growing-season  parameters  were  compared.  Length 
of  the  growing  season  was  indexed  first  as  the  number  of 
months  when  air  temperature  is  above  0  °C,  that  is,  the 
approximate  number  of  months  when  the  soil  is  thawed 


86 


(Weaver  1994).  Second,  drought  months  (none  according 
to  the  Walter  index)  were  subtracted  from  the  warm- 
season  index  to  create  a  better  warm-moist  season  index. 
Despite  contrary  observations  (Weaver  1994),  growth  is 
expected  to  be  indexed  by  the  integral  of  temperature  over 
growing  season.  Thus  two  indices  of  growing  season  pro- 
ductivity (W eaver  1994)  were  calculated:  [growing  season 
X  (average  growing  season  temperature  -5  °C)  (Chang 
1968)]  and,  considering  the  Q    effect,  [growing  season 
X  average  growing  season  Q]. 

WINTER  TEMPERATURES 

Average  winter  (January)  temperatures  in  northeastern 
Asia  are  the  world's  coldest  and  they  warm  southward 
(Fullard  and  Darby  1964).  Normal  daily  lows  in  regions 
occupied  by  the  trees  are  Pinus  pumila  (-30  °C),  P.  sibir- 
ica  (-27  °C),  P.  koraiensis  (-27  °C),  P.  albicaulis  (-14  °C), 
and  P.  cembra  (-8  °C),  respectively  (table  1).  Daily  highs 
in  winter  average  8  °C  higher  (table  1). 

Absolute  lows  experienced  by  Pinus  sibirica  (-55  °C) 
and  P.  pumila  (-52  °C)  rise  to  P.  koraiensis  (-42  °C), 
P.  albicaulis  (-34  °C),  and  P.  cembra  (-23  °C)  (table  1). 
Pinus  albicaulis  and  P.  cembra  may  be  protected  from  ex- 
treme low  temperatures  by  the  drainage  of  cold  air  from 
their  mountaintop  sites.  Becwar  and  Burke  1982  show 
that  timberline  conifers  of  the  Colorado  Rockies  do  not 
survive  temperatures  below  -40  °C;  thus  P.  sibirica  and 
P.  pumila  may  be  the  only  stone  pines  tolerant  of  -50  °C. 

Frosts  of  early  fall  and  late  spring  may  catch  trees  in 
partially  hardened  states.  Absolute  lows  in  the  first  and 
last  months  of  winter  were  -21  °C  for  P.  koraiensis,  -17  °C 
for  P.  sibirica,  and  -10  °C  for  the  remaining  trees.  Why 
are  absolute  lows  for  fall-winter-spring  frosts  lower  in  the 
P.  koriaensis  and  P.  sibirica  regions  than  in  the  P.  pumila 
regions  where  average  temperatures  are  lower  (table  1; 
Fullard  and  Darby  1964)? 


SUMMER  TEMPERATURES 

Continentality  allows  northeastern  Asia  to  warm  in 
summer  more  than  less  continental  parts  of  the  polar 
region.  Thus  absolute  maximum  temperatures  of  Pinus 
sibirica  (37  °C)  and  P.  pumila  (33  °C)  are  considerably 
higher  than  those  of  P.  cembra  (27  °C)  and  P.  albicaulis 
(29  °C)  (table  1).  Absolute  summer  highs  in  the  P.  korai- 
ensis forest  at  the  boreal-deciduous  forest  transition  are 
as  high  (36  °C)  as  in  interior  forests.  Average  July  highs 
in  P.  sibirica  and  P.  koraiensis  forests  are  like  those  in 
Rocky  Mountain  Douglas-fir  ( Pseudotsuga  menziesii )  for- 
ests, a  vegetation  zone  below  the  Rocky  Mountain  stone 
pine  zone  (Weaver  1994). 

Daily  growth  rates  are  determined  by  average  tempera- 
ture conditions  in  the  growing  season.  (1)  Temperature 
conditions  are  represented  most  simply  by  average  tem- 
perature; this  is  13  °C  for  Pinus  sibirica,  11  °C  for  P.  korai- 
ensis, and  8  to  9  °C  for  P.  pumila ,  P.  albicaulis,  and  P.  cem- 
bra. Like  July  highs,  13  °C  average  temperatures,  are 
similar  to  those  found  in  the  Douglas-fir/ponderosa  pine 
forests  of  the  Rocky  Mountains  (Weaver  1994).  (2)  Be- 
cause growth  rises  exponentially  with  increasing  tem- 
perature (Qjq;  Larcher  1975),  a  better  index  of  tempera- 
ture on  growth  may  be  a  cross-season  average  of  growth 
support  units  (Weaver  1994),  which  give  greater  weight 
to  high  than  low  temperatures.  With  this  index  climates 
cool  from  P.  sibirica  to  P.  albicaulis  to  P.  pumila-P.  korai- 
ensis to  P.  cembra  (table  1). 

PRECIPITATION 

Winter-spring  precipitation  is  high  in  Pinus  albicaulis 
(829  mm)  and  P.  cembra  (616  mm)  and  lower  (about 
250  mm)  in  northeastern  Asia  (table  2).  Since  the  lower 
amounts  will  saturate  most  moimtain  soils  (Weaver  1978), 
the  excess  is  expected  to  run  off,  to  have  little  effect  on 
water  supplies  in  forest  stands  during  the  growing  season. 


Table  1 — Temperatures 

^  in  stone  pine  communities  of  the  world. 

Climates  are  listed  in  order  of  increasing  winter  temperatures 

Species,  location,  number  of  stations 

2 

P .  pumila 

P.  sibirica 

P.  koraiensis 

P.  albicaulis 

P.  cembra 

Temperature 

N.  China 

Siberia 

Korea-China 

N.  America 

Euro-Alps 

data 

4+ 

4+ 

3 

4 

3 

Winter  Temperature 

Jan.  mean  min 

-30.0  ±  8.0 

-27.0  ±  2.0 

-27.0  ±  2.0 

-14.0  ±2.0 

-8.0  ±  0.0 

Jan.  mean  max 

-24.0  ±  9.0 

-19.0  ±2.0 

-11.0  ±2.0 

-5.0  ±  3.0 

-1.0  ±  1.0 

Abs  min 

-52.0  ±  4.0 

-55.0  ±  2.0 

-42.0  ±  1.0 

-34.0  ±  2.0 

-23.0  ±  1 .0 

Frost  spring 

-11.0  ±3.0 

-18.0  ±2.0 

-26.4  ±  0.2 

-1 1 .0  ±  1 .0 

-10.0  ±  1.0 

Frost  fall 

-11.0  ±3.0 

-16.0  ±2.0 

-27.6  ±  0.7 

-1 1 .0  ±  1 .0 

-10.0  ±  1.0 

Summer  Temperature 

July  mean  min 

8.0  ±  0.0 

12.0  ±  1.0 

14.0  ±1.0 

4.0  ±  1.0 

5.0  ±  1.0 

July  mean  max 

15.0  ±2.0 

21.0±  1.0 

26.0  ±  0.0 

18.0  ±  1.0 

14.0  ±1.0 

Abs  max 

33.0  ±1.0 

37.0  ±  1.0 

36.0  ±2.0 

29.0  ±  1.0 

27.0  ±  2.0 

9.0  ±  1.0 

13.0  ±1.0 

11.3±  1.2 

9.0  ±1.0 

8.0  ±  1.0 

1.9  ±0.2 

2.4  ±  0.1 

1.9  ±0.1 

2.1  ±0.2 

1.7  ±0.1 

'Temperature  data  (°C)  are  the  mean  ±  one  standard  error.  Absolute  temperatures  are  recorded  for  1 0  years  in  P.  albicaulis  and  P.  cembra; 
records  for  the  Asian  pines  are  unspecified  (presumed  longer).  T^,  and  Q^,  are  growing  season  averages  defined  in  the  text. 

^  Sample  size  is  four  for  P.  albicaulis  (except  average  max  and  min  for  January  and  July,  n  =  5),  three  for  P.  cembra,  and  for  both  P.  pumila  and 
P.  sibirica  eight,  except  for  January  max-min,  July  max-min,  and  spring-fall  frost  temperatures  where  n  =  4. 


87 


and,  thus,  to  have  httle  effect  on  production.  Large  snow- 
falls in  the  P.  albicaulis  and  P.  cembra  forests,  relative 
to  those  in  the  Asian  forests,  surely  result  in  greater 
snowpacks,  which  shelter  ground-level  plants  and  ani- 
mals less  frost  tolerant  than  the  trees. 

Summer  precipitation  is  lower  where  moisture  carrying 
air  masses  cross  mountains  (northeastern  Asia  and  the 
Rocky  Mountains)  than  where  they  do  not  (Alps  and  east- 
ern China).  Thus  summer  rainfall  increases  from  Pinus 
albicaulis  (102  mm)  to  P.  pumila  (187  mm),  P.  sibirica 
(187  mm),  P.  cembra  (323  mm),  and  P.  koraiensis 
(394  mm).  Precipitation  in  the  driest  month  parallels 
summer  rainfall  (table  2). 

Soil  water  may  be  adequate  in  spite  of  low  rainfall  if 
evapotranspiration  is  low.  If  we  assume  that  1  °C  degree 
evaporates  2  mm  per  month  (Nielson  1986,  1992;  Walter 
1973),  we  see  that  stone  pines  do  not  experience  drought 
in  any  month  (table  2).  In  addition,  some  buffering 
against  drought  occurs  because  snowmelt  water  stored  in 
the  soil  provides  a  supplement  to  summer  showers.  How- 
ever, even  at  boreal  temperatures  Walter's  index  may  un- 
derstate evaporation  (Stephenson  1990),  so  there  is  prob- 
ably less  summertime  restriction  of  growth  by  drought 
(stomate  closure)  in  Pinus  cembra  and  P.  koraiensis  than 
in  P.  albicaulis,  P.  sibirica,  and  P.  pumila  forests. 

Organisms  without  water  reserves — such  as  lichens, 
mosses,  and  invertebrates — are  more  sensitive  to  summer 
drought  and  may  distinguish  wetter  and  drier  forests. 
These  organisms  are  surely  inhibited  in  Pinus  pumila 
and  P.  albicaulis  forests  (one  to  eight  rain  days/month) 
relative  to  P.  cembra  and  P.  koraiensis  forests  (16  to  18 
rain  days). 

GROWING  SEASON 

Survival  depends  on  photos3aithetic  provisioning  for 
winter  respiration  and  the  outcome  of  competition  may 
depend  on  excesses  above  this  basic  provisioning.  One 
expects  production  to  be  correlated  with  the  number  of 
warm-moist  days,  with  the  warmth  of  those  days  (  Q^^, 
Larcher  1975),  and  thus  with  their  product. 


If  water  stress  never  occurs,  the  length  of  the  warm- 
moist  season  increases  from  Pinus  albicaulis  (4.5  mo) 
to  P.  pumila  (4.6  mo),  P.  sibirica  (5.5  mo),  P.  cembra 
(6.3  mo),  and  P.  koraiensis  (7.8  mo)  (table  3).  And  if  tem- 
perature conditions  were  identical  in  these  forests,  one 
would  therefore  expect  production  in  P.  cembra  and 
P.  koraiensis  forests  to  be  half  again  as  great  as  in 
P.  pumila  and  P.  albicaulis  forests.  Actual  differences 
could  be  even  larger  because  the  temperature-defined 
growing  season  may  overstate  stand  productivity  in  drier 
regions.  For  example,  while  subsoils  remain  moist  all 
summer  in  higher  (J.  Brown,  personal  commimication) 
and  lower  (Weaver  1974)  parts  of  the  P.  albicaulis  zone, 
drying  of  surface  soils  apparently  causes  tree  water  stress 
at  lower  (B.  Keane,  personal  communication),  but  not 
higher  (J.  Brown,  personal  communication),  sites.  Similar 
droughty  periods  probably  occur  in  warmer  P.  pumila 
and  P.  sibirica  forests. 

Production  is  expected  to  rise  exponentially  with  rising 
temperature,  and  thus  productivity  might  be  better  corre- 
lated with  the  product  of  season  length  and  temperature 
"growth  supporting  units."  While  this  production  hy- 
pothesis was  rejected  in  a  cross-vegetation  zone  analysis 
(Weaver  1994),  the  genetic  similarity  of  stone  pines  might 
allow  it  to  operate  here.  If  so,  and  if  water  deficits  in 
Siberia  do  not  shorten  the  growing  season,  P.  koraiensis 
and  P.  sibirica  will  be  promoted  to  the  most  productive 
stone  pines  (table  3). 

CONCLUSIONS 

From  a  tree's  point  of  view,  average  conditions  in  the 
stone  pine  zones  are  similar:  water  stress  is  slight  or 
nonexistent  and  growing  season  temperatures  average 
9  to  13  °C.  Seasonal  extreme  conditions  may,  however, 
prevent  reciprocal  transplantation.  Winter  or  fall-spring 
frosts  in  the  Pinus  sibirica/P.  pumila/P.  koraiensis  region 
may  exclude  the  other  pines.  And  weak  droughts  in  the 
P.  albicaulis/P.  pumila/P.  sibirica  regions  might  exclude 
P.  cembra  or  P.  koraiensis. 


Table  2 — Precipitation^  in  stone  pine  forests  of  ttie  world.  Climates  are  listed  in  order  of  increasing  summer  precipitation 


Species,  location,  number  of  stations  ^ 

P .  albicaulis 

P.  pumila 

P.  s  ib  irica 

P.  cembra 

P.  koraiensis 

Precipitation 

N.  Annerica 

N.  China 

Siberia 

Euro-Alps 

Korea-China 

data 

4 

4+ 

4+ 

3 

3 

Total 

931  ±  229 

407  ±137 

432  ±  21 

939  ±  9 

636  ±  68 

Oct.-June 

829  ±  234 

264 ±  76 

245  ±  22 

616  ±43 

242  ±  62 

July-Sept. 

102  ±  14 

143  ±  43 

187  ±11 

323  ±  36 

394  ±  17 

Wettest  summer  month 

116±  16 

165 ±  25 

181  ±  8 

214±  15 

346 

Driest  summer  month 

4±  4 

4±  2 

8±  2 

45  ±16 

32 

Summer  drought  months 

0±  0 

0±  0 

0±  0 

0±  0 

0±  0 

Summer  water  deficit 

0±  0 

0±  0 

0±  0 

0±  0 

0±  0 

Summer  rain  days, 

average  number 

8±  1 

1  ±  1 

14±  1 

16±  1 

18 

'Precipitation  (mm)  data  are  total  (sum  of  all  months),  winter,  summer  (July,  August,  plus  September),  driest  summer  month  (July-September)  recorded, 
wettest  summer  month  (July-September)  recorded,  and  average  number  of  showers  in  June-September.  High  variances  in  total  precipitation  for  P. 
albicaulis  and  P.  pumila  are  reduced  to  705  ±  51  mm  and  274  ±  37  mm  by  omission  of  the  Crater  Lake  and  Petropavlovsk  stations,  respectively. 

^Sample  size  is  four  for  P.  albicaulis,  three  for  P.  cembra,  three  for  P.  Koraiensis,  and  eight  for  both  P.  pumila  and  P.  sibirica,  except  for  summer  drought 
months  and  deficit,  where  it  is  only  four. 


88 


Table  3 — Stone  pine  productivity  correlates:  growing  season  ,^  growing  season  temperature,^  and  their  products.  Climates  are  listed  in 
descending  order  by  one  estimate  of  productivity,  warm  moist  season 

 Species,  location,  number  of  stations  

P.  koralensis  P .  cembra  P.  siblrica  P.  pumlla  P.  albicaulis 

Korea-China  Euro-Alps  Siberia  N.  China  N.  America 


4 

4 

4 

Warm  season 

7.8  ±0.1 

6.3  ±  0.2 

5.5  ±  0.2 

4.6  ±  0.4 

4.5  ±  0.3 

Summer  drought  months 

0.0  ±0.0 

0.0  ±  0.0 

0.0  ±0.0 

0.0  ±0.0 

0.0  ±0.0 

Warm-moist  season 

7.8  ±0.1 

6.3  ±  0.2 

5.5  ±  0.2 

4.6  ±  0.5 

4.5  ±  0.3 

k 

GS  X  Tg -5  °C 
GSxQg, 

11.3±1.2 

8.0  ±1.0 

13.0  ±  1.0 

9.0  ±  1.0 

9.0  ±  1.0 

1.9  ±0.1 

1.7  ±0.1 

2.4  ±0.1 

1.9  ±0.2 

2.1  ±0.2 

56.5 

18.9 

44.0 

18.4 

18.0 

15.6 

10.7 

13.2 

8.7 

9.5 

'Warm-season  and  warm-moist  season  are  two  indices  of  growing  season.  Warm-season  months  occur  after  average  monthly  air  temperature  hses 
above  0  °C  and  before  it  falls  below  0  °C.  The  warm-moist  season  Is  the  warm  season  minus  any  months  in  which  T/2  is  greater  than  P(mm),  Walter  1973; 
"Walter  drought"  does  not  occur  in  stone  pine  regions. 

H'wo  indices  of  growing-season  temperature  are  provided.  T^,  is  the  average  temperature  in  growing  season  months.  T  -5C  is  used  as  one  index  of 
growth  support  on  the  assumption  that  growth  does  not  occur  below  5  °C  (Chang  1 968).  Q  is  the  average  of  temperatures  weighted  tor  their  growth- 
supporting  capacity:  0  °C  =  0,  1  °C  =  1 ,  1 1  °C  =  2,  21  °C  =  4  {Weaver  1993). 


Organisms  filling  other  niches  might  find  the  climates 
dissimilar  (Weaver  1990).  Poikilohydric  organisms  and 
himaans  would  react  differently  to  the  relatively  smnmer- 
rainy  climates  of  the  P.  cembra  and  P.  koraiensis  regions 
than  to  the  other  stone  pine  climates.  And  organisms 
wintering  imder  snow — small  mammals,  insects,  plants 
of  small  stature — are  more  protected  from  the  extremes 
of  winter  cold  in  the  P.  cembra  and  P.  albicaulis  regions 
than  in  the  other  climates. 

ACKNOWLEDGMENTS 

I  thank  (1)  W.  Tranqmllini  (University  of  Innsbruck, 
Austria)  for  providing  data  from  his  P.  cembra  weather 
stations,  (2)  J.  Luo  (Beijing  Forestry  University,  China) 
and  Z.  Shidong  (Inst.  Appl.  Ecol.,  Shenyang,  China)  for 
providing  data  from  weather  stations  in  P.  koraiensis  cli- 
mates, (3)  H.  Lieth  (University  of  Osabuck,  Germany), 
P.  Gorchakovsky  (Acad.  Sci.  Inst.  Ecol.,  Ekaterinburg, 
Russia)  and  P.  Khomentovsky  (Kamchatka  Inst.  Ecol., 
Petropavlovsk,  Russia)  for  help  in  selecting  P.  sibirica 
and  P.  pumila  weather  stations,  and  (4)  nineteen  North 
American  ecologists  already  acknowledged  (Weaver  1990) 
for  help  in  selecting  North  American  P.  albicaulis  stations. 
I  am  also  grateful  for  support  to  the  project  provided 
through  Intermountain  Research  Station  contract  #INT- 
92720-RJVA  (W.  Schmidt). 

REFERENCES 

Becwar,  M.;  Burke,  M.  1982.  Winter  hardiness  limitations 

and  physiography  of  woody  timberline  flora.  In:  Li,  P.; 

Sakai,  A.  Plant  cold  hardiness  and  freezing  stress.  New 

York:  Academic  Press:  307-324. 
Chang,  J.  1968.  Climate  and  agriculture.  Chicago:  Aldine. 

304  p. 

Fullard,  H.;  Darby,  H.  1964.  The  imiversity  atlas.  London: 
George  Philip  and  Son.  176  p. 

Lanner,  R.  1990.  Biology,  taxonomy,  evolution,  and  ge- 
ography of  stone  pines  of  the  world.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 


whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Larcher,  W.  1975.  Physiological  plant  ecology.  New  York: 
Springer.  252  p. 

Mirov,  N.  1967.  The  genus  Pinus.  New  York:  Ronald 
Press.  602  p. 

Muller,  M.  1982.  Selected  climatic  data  for  a  global  set 
of  standard  stations  for  vegetation  science.  The  Hague, 
Netherlands:  Junk.  306  p. 

Neilson,  R.  1986.  High  resolution  climatic  analysis  and 
southwest  biogeography.  Science.  232:  27-33. 

Nielsen,  R.  1992.  Toward  a  rule  based  biome  model.  Land- 
scape Ecology.  7:  27-43. 

Stephenson,  N.  1990.  Climatic  control  of  vegetation  distri- 
bution: the  role  of  water  balance.  American  Naturalist. 
135:  649-670. 

Walter,  H.  1973.  Vegetation  of  the  earth  in  relation  to  cli- 
mate and  ecophysiolgical  conditions.  New  York: 
Springer.  237  p. 

Weaver,  T.  1974.  Root  distribution  and  soil  water  regimes 
in  nine  habitat  types  of  the  northern  Rocky  Mountains. 
In:  Marshall,  J.,  ed.  The  belowgroimd  ecosystem.  Sci. 
Ser.  26.  Fort  ColHns,  CO:  Colorado  State  University, 
Range  Science  Department.  351  p. 

Weaver,  T.  1979.  Changes  in  soils  along  a  vegetational 
(altitudinal)  gradient  of  the  northern  Rocky  Mountains. 
In:  Youngberg,  C,  ed.  Proceedings  of  the  Fifth  North 
American  forest  soils  conference.  Madison,  WI:  Soil  Sci- 
ence Society  America.  14-29. 

Weaver,  T.  1990.  Climates  of  subalpine  pine  woodlands. 
In:  Schmidt,  W.;  McDonald,  K.,  comps.  Proceedings — 
s)rmposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-moimtain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
72-79. 

Weaver,  T.  1994.  Vegetation  distribution  and  production 
in  Rocky  Mountain  climates — wdth  emphasis  on 
whitebark  pine  [these  proceedings]. 


89 


Growth  Characteristics 


International  Workshop 
St.  Moritz  1 992 


90 


GROWTH  OF  SWISS  STONE  PINES 
THAT  ORIGINATED  FROM  AND  WERE 
PLANTED  AT  SEVERAL  ALTITUDES  IN 
THE  AUSTRIAN  ALPS 

Kurt  Holzer 


Abstract — For  the  purpose  of  gene  conservation,  a  plus  tree 
collection  of  Swiss  stone  pine  {Pinus  cembra)  was  conducted. 
Scions  of  185  trees  and  82  open-pollinated  seed  samples  were 
collected  within  the  natural  range  of  Swiss  stone  pine  in  Austria, 
In  1987,  clonal  height  was  measured  at  an  age  between  25  and 
30  years.  Open-pollinated  families  were  planted  at  three  differ- 
ent sites.  Height  was  measured  after  18  years.  Growth  of  graft- 
ings and  open-pollinated  families  helped  identify  the  altitude  of 
their  origin.  When  planted  below  timberline,  growth  decreased 
with  increasing  altitude. 


In  1957,  the  Department  of  Forest  Tree  Breeding  and 
Genetics,  Federal  Forestry  Research  Institute,  Austria, 
commenced  a  selection  program  of  Swiss  stone  pine  (Pi- 
nus cembra).  The  objectives  of  this  program  were  to 
preserve  genetic  resovirces  of  superior  trees  in  a  clonal 
orchard  and  to  assess  the  growth  capacity  of  the  clones 
by  field  performance  of  open-pollinated  families. 

Within  the  natural  range  of  stone  pine,  timberline  di- 
vides the  distribution  of  stone  pine,  due  to  environmental 
conditions,  into  two  silvicultural  zones:  (1)  production  for- 
ests and  (2)  "Kampfzone."  Hence,  those  aspects  had  to  be 
considered  for  the  program  (Holzer  1963, 1976). 

MATERIAL  AND  METHODS 

Covering  the  majority  of  natural  stands  in  Austria, 
185  superior  trees  were  selected  between  1958  and  1965 
(Holzer  1961,  1969).  Within  production  forests  scions  of 
124  and  above  timberline  ("Kampfzone")  scions  of  61  indi- 
viduals were  collected.  The  altitudinal  range  was  approxi- 
mately 250  m  below  and  230  m  above  timberline  (be- 
tween 1,650  m  and  2,200  m  above  sea  level).  In  addition, 
open-pollinated  seeds  were  harvested  from  82  trees. 

Graftings  were  planted  in  a  conservation  orchard  in 
Purkersdorf-Stadlhutte  near  Vienna  at  400  m  above  sea 
level.  Progenies  were  raised  in  the  nursery  of  the  insti- 
tute (Mariabrvmn  Vienna,  220  m  above  sea  level)  for 
6  to  8  years  and  then  planted  at  three  different  sites: 
(1)  Purkersdorf-Stadlhutte;  (2)  GroBe  Zirbenwiese  (1,650  m 
above  sea  level);  Seetaler  Alpen,  a  swampy  meadow 
within  the  natiiral  range  of  stone  pine,  approximately 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Kurt  Holzer  is  Professor  of  Forest  Genetics,  Institut  fur  Waldbau, 
Universitat  fur  Bodenkultur,  A-1190  Vienna,  Austria. 


250  m  below  timberline;  and  (3)  Melcheben,  Packalpe 
(1,700  m  above  sea  level),  close  to  timberline  of  Norway 
spruce  (Picea  abies).  At  sites  2  and  3,  only  32  families 
were  planted  because  of  limited  nvmabers  of  plants. 

In  1987,  height  was  measvired  at  a  clonal  age  between 
25  and  30  years.  Heights  of  progenies  were  assessed  at 
an  age  of  8  and  18  years,  respectively.  Exclusively  de- 
scriptive statistics  were  employed. 

Some  results  have  been  previously  published  with 
respect  to  graftings  (Holzer  1989)  and  progenies  (Holzer 
1978).  Recently,  growth  of  vegetative  and  generative 
plant  material  has  been  measured  (Feuersinger  1992). 

Since  altitude  of  timberline  is  irregular  within  the 
Alps,  comparisons  are  based  on  relative  distances  to  local 
timberline. 

RESULTS  AND  DISCUSSION 

Results  are  summarized  in  figures  1-3  within  the  main 
distribution  (about  150  m  below  to  100  m  above  timber- 
line);  mean  height  of  30-year-old  graftings  decreased  with 
increasing  elevation  of  origin  fi-om  3.8  m  at  150  m  below 
to  2.7  m  at  100  m  above  timberline.  Variation  was  high 
and  yielded  to  50  to  70  percent  of  the  clonal  mean.  In 
general,  the  majority  of  graftings  originating  below  tim- 
berline had  superior  growth.  Presimiably,  inherent 
growth  capacity  shows  a  cline  within  the  main  distribu- 
tion zone  of  stone  pine  (fig.  1). 

This  growth  pattern  was  also  found  in  generative  plant 
material.  Progenies  originating  fi'om  100  m  below  timber- 
line  showed  pronoimced  reduced  growth.  However,  when 
growth  was  only  evaluated  within  a  single  seed  zone,  it 
was  always  higher  below  the  timberline  than  above  (fig.  2). 
In  the  nursery,  height  of  families  (n  =  50)  originating  fi'om 
production  forests  averaged  46.5  cm.  Families  in  =  16)  fi'om 
the  "Kampfzone"  had  a  slightly  reduced  growth  of  40.0  cm 
at  an  age  of  8  years. 

Transplants  of  families  18  years  old  showed  different 
results.  Clinal  growth  pattern  is  still  distinguishable  at 
site  1  (400  m  above  sea  level)  and  site  2  (250  m  below  tim- 
berline). However,  at  site  3  (timberline)  height  growth  re- 
actions were  different.  Here,  families  with  small  height 
growth  potential  at  400  m  above  sea  level  were  superior. 
Figure  3  shows  growth  of  families  originating  fi'om  two 
different  seed  zones.  In  seed  zone  "Defi-eggen"  growth 
of  18  families,  400  m  and  1,700  m  above  sea  level,  was 
negatively  correlated  (r  =  -0.752***).  Fovuteen  families 
originating  from  seed  zone  "Prankerhohe"  showed  a  con- 
sistent pattern.  However,  correlation  (r  =  -0.205)  was  not 
significant. 


91 


200 -X 

150- 

100- 

? 

£  50- 

O) 

O  above 

f timber  line 
below 

50  - 
100- 

150- 
200- 
250- 


o 

3 


O  X  X  J     X  X  S 

I 


fix  X 


-|^T<  5-5-xW^X-X-^-> 


XXX  X.J 

X  X  Sx  X  S  X  X  X  >j6^  I  i  ^ 

N 

XX  Jgx      X^xg)^  X 

I  \ 

X  xx8xi>x'*8    X  X 

I  X 

/ 

X       X  I     XX  X 

X  S  I   X  P         XX  X 

/ 

l/ 

I  I  I 
3  4 

Height  of  Graftings  (m) 


Figure  1 — Total  height  growth  of  30-year-olcl 
graftings  In  the  plantation  at  400  m  above  sea 
level  with  the  altitude  of  their  origin  (distance 
from  the  local  timberline).  Dotted  curve  denotes 
the  mean  values. 


130 


120 


110  - 


100 


90  - 


1 ,700  m  asl 

X 


o  o 


X  \ 

x"^  X 

X  \ 

\ 


X 

o 


\ 


400  m  asl 


100 


120  140 

Height  (cm) 


160 


Figure  3 — The  comparison  of  the  mean  of  the 
1 8-year-old  families  on  two  different  planting 
sites:  400  m  and  1 ,700  m  above  sea  level,  re- 
spectively (provenance  1/6/100,  Prankerhohe, 
marked  with  o,  provenance  1/8/100,  Defreggen, 
marked  with  x;  see  text  also).  Curve  denotes 
correlation  (r  =  -0.752)  of  the  Defreggen  zone. 


This  observation  is  supported  by  a  sowing  trial  200  m 
above  local  timberline  (2,100  m  above  sea  level).  Pro- 
genies originating  from  sites  above  timberline  performed 
well;  progenies  from  trees  in  production  forests  performed 
poorly  (Holzer  1975). 


What  conclusions  regarding  growth  of  stone  pine  might 
be  drawn  from  these  experiments  so  far? 

1.  Height  growth  of  stone  pine  shows  clinal  variation. 

2.  When  plants  are  grown  in  production  forests,  growth 
is  positively  correlated  to  elevation. 

3.  Above  timberline  this  correlation  is  negative. 


200- 
150- 
100- 


•^50-  e 
c 

S above 
timber  line 
o  below 


•E  50- 
< 

100- 


4Q  0 

+ 


-X-XQW-^ 


1s 


o  +        ++  oma  _pco>ooo  0    o    o  o 


+  +  +  Ot  + 


150- 


200- 


■H-X  |X         X"xp        X  XX 


30 


-H-O-l-  +■ 

I 

40 


50  60 

Height  of  Progenies  (cm) 


70 


Figure  2 — Total  height  of  the  8-year-old  prog- 
enies (mean  of  families)  of  the  harvested 
clones  with  the  altitude  of  their  origin  (distance 
from  the  local  timberline).  Dotted  curve  de- 
notes the  mean  values. 


REFERENCES 

Feuersinger,  P.  1992.  Das  Wachstum  von  Zirbe  in 
Abhangigkeit  vom  Ursprungsort.  Diplomarbeit,  Univ.  f. 
Bodenkultur,  Wien.  140  p. 

Holzer,  K.  1961.  Der  derzeitige  Stand  der  Auswahl  von 
Schwarzerlen  und  von  Zirbe  im  Rahmen  der 
Forstpflanzenziichtung.  Informationsdienst  50,  Allg. 
Forstztg.  72. 

Holzer,  K.  1963.  Pinus  cembra  L.  as  a  pioneer  at  timber- 
line  in  the  European  Alps.  Proc.  World.  Cons.  Forest 
Genet,  and  Tree  Imprv.,  Vol.  1  (FAO/FORGEN  1963, 
3/13,  FAO,  Rome). 

Holzer,  K.  1969.  Erste  Ergebnisse  von  Zirbenauswahl- 
baumen  {Pinus  cembra  L.).  Centrbl.f.d.  ges.  Forstwes. 
86:  149-160. 

Holzer,  K.  1975.  Genetics  of  Pinus  cembra  L.  Annales 

Forestales,  Zagreb.  6(5):  139-158. 
Holzer,  K.  1976.  Breeding  haploxylon  pines  for  subalpine 

regions.  XVI  lUFRO-Congr.  Oslo.  Div.  II:  216-227. 
Holzer,  K.  1978.  Versuchsanlage  GroBe  Zirbenwiese.  Exk. 

Fiihrer  Arbeitsgem.  f.  Hochlagenaufforstung.  5  p. 

[Mimeo.] 

Holzer,  K.  1989.  Drei  Jahrzehnte  Erfahrungen  mit 
Zirbenpfropfimgen.  Centrbl.  f.d.  ges.  Forstwes.  106: 
79-88. 


92 


SEASONAL  PATTERNS  OF  GROWTH 
AND  PHOTOSYNTHETIC  ACTIVITY  OF 
PINUS  PUMILA  GROWING  ON  THE  KISO 
MOUNTAIN  RANGE,  CENTRAL  JAPAN 

Takuya  Kajimoto 


Abstract — Seasonal  patterns  of  growth  and  photosynthesis  in 
Pinus  pumila  needles  were  investigated  in  relation  to  carbon 
uptake.  Current  needles  elongated  from  July  to  late  August. 
Photosynthetic  activities  of  2-year-old  needles,  both  in  sun-  and 
shade-needles,  reached  seasonal  maxima  in  August,  while  those 
of  current  needles  did  so  in  September.  The  results  suggest  that 
annual  net  carbon  gain  of  P.  pumila  depended  mainly  on  net  pho- 
tosynthesis of  old  needles  (1  to  4  year  olds). 


Growth  and  production  rates  of  trees  within  the  timber- 
Une  ecotone  are  generally  restricted  mainly  due  to  the 
short  growing  season  (Tranquillini  1979).  However,  most 
of  the  subalpine  species  show  physiological  adaptations  to 
conditions  associated  with  low  temperature,  lower  photo- 
synthetic  optimun  temperature  (Larcher  1975),  and 
higher  frost  and  freezing  resistance  (Sakai  and  Larcher 
1987).  It  has  also  been  suggested  that  within  krummholz 
mats  consisting  of  Picea  engelmannii  and  Abies  lasiocarpa 
photosynthesis  is  enhanced  by  more  favorable  canopy- 
temperature  conditions  (Hadley  and  Smith  1987).  Infor- 
mation on  such  physiological  and  structural  functions  is 
indispensable  for  understanding  how  alpine  trees  survive 
by  achieving  positive  carbon  balance  during  the  short 
siunmers. 

In  Japan,  Pinus  pumila  Kegel,  occurs  in  high-moimtain 
areas  from  central  Honshu  (the  main  island)  to  the  north- 
em  island  of  Hokkaido.  This  pine  generally  exhibits 
dwarfed  growth  forms,  and  regenerates  by  layering  at 
the  mature  growth  stage  (Kajimoto  1992;  Okitsu  and  Ito 
1984).  The  physiognomy  of  P.  pumila  is  similar  to  Pinus 
mugo  at  the  upper  timberline  in  the  European  Alps  or  in 
the  Carpathian  Mountains  (Holtmeier  1973,  1981;  Wardle 
1977). 

The  foliage  biomass  of  P.  pumila  rsinges  from  15  to  24 
tons  d.w./ha  despite  the  low  tree  heights.  Pinus  pumila 
stands  develop  dense  canopies  with  leaf  area  density  of 
about  5  mVm^,  which  is  considerably  more  than  in  other 
conifer  forests  (Kajimoto  1989a).  This  canopy  structure 
may  result  in  large  annual  net  production  (Okitsu  and  Ito 
1989;  Shidei  1963).  However,  seasonal  changes  in  growth 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Takuya  Kajimoto  is  a  special  Researcher  of  Science  and  Technology, 
Laboratory  of  Plant  Production,  Forestry  and  Forest  Products  Research 
Institute,  Tsukuba,  Ibaraki  305,  Japan. 


and  photosynthesis  of  P.  pumila  needles  are  still  un- 
known in  relation  to  their  carbon  uptake. 

In  this  paper,  a  process  of  photosynthetic  production 
in  P.  pumila  during  the  growing  season  is  discussed 
based  on  observations  of  environmental  factors  and 
seasonal  patterns  of  growth  and  photosynthesis  in  the 
needles. 

STUDY  AREA 

The  study  area  is  located  on  the  Kiso  Mountain  Range 
in  central  Japan.  The  range  trends  from  south  to  north. 
The  highest  peak  is  Mount  Kisokoma  (2,965  m  a.s.l.). 
In  this  mountain  range,  P.  pumila  trees  naturally  occur 
above  2,500  m.  The  study  was  carried  out  in  the 
Shinshu  University  Experimental  Forest  (35°48'  N., 
137°50'  E.)  in  the  northern  part  of  the  Kiso  Mountain 
Range.  The  investigated  pine  stand  is  located  on  the 
northeast  slope  (15°  inclination)  at  an  altitude  of 
2,600  m  a.s.l.  Dense  canopy  layer  is  developed  between 
100  and  200  cm  above  the  ground  height  (Kajimoto 
1989a). 

FIELD  OBSERVATIONS 

Four  siin-exposed  terminal  shoots  of  P.  pumila  were 
selected.  Lengths  of  current  shoots,  current  needles, 
and  winter  buds  were  measured  seven  times  between 
late  May  and  October  1988.  Current  needle  length  of 
each  shoot  was  determined  as  the  average  of  10  fascicles 
selected  at  equidistant  intervals  along  the  shoot  axis. 

At  the  plot,  solar  irradiance  (400-1,100  nm  in  wave- 
length) above  the  canopy  siuface  was  measiu-ed  with  a 
pyranometer  sensor  (LICOR,  LI-200SB)  at  1-hour  inter- 
vals from  June  to  October  1988.  Photos3nithetic  photon 
fliix  density  (PFD)  (400-700  nm)  above  the  canopy  sur- 
face and  air  temperature  inside  the  stand  were  also 
recorded  at  1-hour  intervals  between  mid-July  and 
October  1988  using  a  quantum  sensor  (Koito,  IKS-25) 
and  a  resistance  thermometer  (Koito,  OPT-150). 

MEASUREMENT  OF 
PHOTOSYNTHESIS 

Rates  of  COg  exchange  were  determined  on  detached 
shoots  that  were  collected  50  hours  prior  to  the  meas- 
urement. The  shoots  were  carefully  transported  to  the 
laboratory  to  minimize  error  associated  with  excision 
and  transportation  (Kajimoto  1990).  An  open  system 


93 


with  an  infrared  gas  analyzer  (Hartmann  and  Braun, 
URA-2)  was  used.  The  air  was  fed  into  an  assimilation 
chamber  (9  by  25  by  1  cm^)  at  a  rate  of  O.Sl/min.  COg  con- 
centration of  the  input  air  was  within  the  range  of  360- 
400  ppm.  The  chamber  was  placed  in  a  thermoregulated 
waterbath  and  illimiinated  with  incandescent  lamps. 
PFD  at  the  chamber  surface  was  measured  with  a  quan- 
tum sensor  (LICOR,  LI-190SB).  Details  of  the  system 
were  described  by  Kajimoto  (1990). 

Two  terminal  shoots  (about  10  cm  in  length)  of  P. 
pumila  were  sampled  from  the  sun-exposed  canopy  sur- 
face part  (1.8-2.0  m)  and  in  the  lower  shaded  canopy  lay- 
er (1.0-1.2  m)  seven  times  between  late  May  and  October 
1988.  At  each  sampling  time,  about  26  fascicles  of  ciirrent 
needles  (flushed  in  1988)  and  2-year-old  needles  (flushed 
in  1986)  were  separately  removed  from  each  sun-  and 
shade-shoot.  The  needles  were  placed  in  the  assimilation 
chamber  keeping  the  cut  end  of  each  fascicle  immersed  in 
water  using  a  small  vinyl  tube.  Net  photosynthetic  rates 
at  different  PFD  levels  (20-1,300  ^imol/mVs)  and  dark  res- 
piration rates  of  these  two  needle  groups  were  measured. 
Air  temperature  inside  the  chamber  was  maintained  at 
10  °C  throughout  the  measurement;  this  temperature 
value  was  within  the  range  of  optimum  air  temperatiire 
(10-15  °C)  in  net  photosynthesis  for  P.  pumila  needles 
(Kajimoto  1990). 


25 


20  - 


15 


o 


<  0 

-5 

-10 
25 


max. 


c 
o 


20  - 


1^ 


10 


£^  5 
o 

^  0 


M 


J  A 
1988 


0  N 


Figure  1 — Seasonal  changes  in  daily  solar  irradiance 
(400-1 ,100  nm  in  wavelength)  and  air  temperatures 
during  1988  measured  in  a  P.  pumila  stand  (2,600  m 
a.s.l.),  Kiso  Mountain  Range.  Each  point  stands  for 
the  mean  value  at  1  -week  intervals. 


£ 

J3 


V) 

■*-' 
C 
Ct 

V. 

O 

o 

U) 


C 

o 

Q> 

"8 

c 


—I  O 


Figure  2 — Seasonal  growth  patterns  of  current 
shoots,  current  needles,  and  winter  buds  ob- 
served for  sun-exposed  P.  pumila  shoots.  Each 
point  shows  the  mean  for  n  =  4  shoots,  o  = 
shoot  length;  •  =  needle  length;  □  =  bud 
length. 


Needle  area  of  each  sample  was  measured  as  its  pro- 
jected area  with  an  area  meter  (Hayashi,  AAC-100),  and 
its  dry  weight  was  measured  after  ovendrying  at  85  °C. 
Specific  leaf  area  (SLA;  cm7g)  was  determined  as  the  ra- 
tio of  needle  area  to  dry  weight.  Net  assimilation  rates 
obtained  in  all  measurements  were  converted  to  the  val- 
ues at  a  COg  concentration  of  380  ppm. 

RESULTS 

Climatic  Conditions 

Daily  solar  radiation  at  the  P.  pumila  stand  was  higher 
in  June  and  gradually  decreased  until  October  in  1988 
(fig.  1).  Mean  air  temperature  was  about  12  °C  during 
August  and  fell  below  0  °C  after  mid-October.  Monthly 
precipitation  was  212  mm  in  July  and  167  mm  in  August 
1988,  according  to  the  data  obtained  at  Senjoujiki  Station 
(2,623  m  a.s.l.)  in  the  central  part  of  Kiso  Mountain  Range 
(Nagano  Meteorological  Observatory  1988). 

Growth  Patterns  of  Shoots  and 
Needles 

The  current  shoots  elongated  rapidly  from  June  to  July 
(fig.  2).  The  current  needles  began  to  flush  in  early  July 
and  elongated  rapidly  during  August.  In  late  May,  the 
winter  buds,  which  would  flush  in  the  next  season,  were 
visible.  The  length  of  winter  bud  increased  gradually  un- 
til late  October. 

Specific  leaf  area  (SLA)  of  the  current  needles,  both 
sun-  and  shade-needles,  was  quite  large  in  early  August 
when  the  needles  began  to  grow  larger,  and  sharply  de- 
creased by  mid-September  (fig.  3).  SLA  of  2-year-old 


94 


Cur -Shade 


Figure  3 — Seasonal  changes  in  specific  leaf  area 
(SLA)  of  the  P.  pumila  needles  used  for  photosyn- 
thesis measurements.  □  =  current,  sun-needles; 
■  =  current,  shade-needles;  O  =  2-year-old,  sun- 
needles;*  =  2-year-old,  shade-needles. 


needles  became  slightly  larger  in  August  than  in  other 
months.  The  shade-needles,  both  current  and  2-year-old, 
showed  larger  SLA  values  throughout  the  season  than 
sun-needles. 


Cur.- Sun 


/ 


Cur-Shod  e 
2  yr-Sun 


2yr-Shade 


1000  1200 


Photosynthetic  photon  flux  density 
(pmol  nn-2s~i ) 

Figure  A — Examples  of  the  relationships  between 
photosynthetic  photon  flux  density  (PFD)  and  net 
photosynthetic  rate  (PJ  at  10  °C  of  P.  pumila 
needles  on  September  15, 1988.  □  =  current,  sun- 
needles;  ■  =  current,  shade-needles;  0=  2-year-old, 
sun-needles;  •  =  2-year-old,  shade-needles. 


PFD 


Light  Response  of  Net  Photosynthesis 

Light  response  of  photosynthesis  was  examined  based 
on  the  relationships  between  photon  flux  density  (PFD) 
and  net  photosynthetic  rates  (PJ  at  10  °C  (fig.  4).  Light 
saturation  points  of    were  different  between  sun-  and 
shade-needles,  and  also  between  needles  of  different  age. 
However,  all  sampled  needles  were  completely  light  satu- 
rated in     at  PFD  value  below  1,000  |imol/m7s  and  10  °C. 
A  maximum  net  photos5nnthetic  rate  (P^^)  was  defined  as 
P^  value  at  1,000  |imol/mVs  and  10  °C,  and  an  approxi- 
mate light-saturation  point  in  net  photosynthesis  (PFDg^) 
was  also  determined  as  PFD  value  that  gave  90  percent 
ofP  . 

max 

Table  1  shows  that  PFDgg  of  2-year-old  needles  ranged 
from  180  to  500  |imol/mVs  in  sun-needles  and  180  to 
440  |imol/m7s  in  shade-needles.  The  values  were  rela- 
tively lower  than  those  of  the  current  needles,  where 
PFDgo  was  440-580  [imol/mVs  in  sun-needles  and  300-500 
\imo]/mVs  in  shade-needles.  Both  current  and  2-year-old 
needles  had  highest  PFDgg  values  in  August  compared  to 
the  other  months. 

Seasonal  Change  in  Photosynthetic 
Activity 

Figure  5  shows  that  P^^  of  current  needles  were  larger 
in  sun-needles  than  in  shade-needles  for  each  period.  For 
2-year-old  needles,  there  was  a  remarkable  difference  in 

between  sun-  and  shade-needles  in  July  and  August. 
The  values  of  sun-  and  shade-needles  both  became  higher 
in  August  than  in  other  months.  P    values  of  current 

^  max 

needles  increased  between  late  August  and  mid-September. 

Dark  respiration  rates  at  10  °C  (R)  of  current  needles 
were  larger  in  sun-needles  than  in  shade-needles,  while 
there  were  no  differences  in  2-year-old  sun-  and  shade- 
needles  (fig.  5).  Current  needles  showed  larger  R  values 
than  2-year-old  needles  throughout  the  season. 


Table  1 — Approximate  light-saturation  points  (PFDgg)  in  the  net 
photosynthetic  rates  (at  1 0  °C)  of  P.  pumila  needles  in 
1988 

Light-saturation  points  (PFD^Y 


Date 

Current  needles 

2-year-old-needles 

Sun 

Shade 

Sun 

Shade 

May  27 

420 

440 

July  2 

340 

340 

July  20 

300 

300 

August  4 

500 

500 

500 

380 

August  24 

580 

460 

480 

380 

September  1 5 

440 

360 

180 

180 

October  20 

500 

300 

240 

220 

'PFDgg  was  defined  as  the  value  of  photosynthetic  photon  flux  density 
which  gave  90  percent  of  a  maximum  net  photosynthetic  rate  {P^^J-  Each 
PFDg^  value  was  determined  using  the  light-photosynthetic  curve,  as  shown  in 
figure  4. 


95 


«»  12 

«A 

*  ft, 

c  o 
w   o  6 

o  e 

Q. 

<U  _  £  ^ 
0 

<u 

■4-^  MIC 

a  V  1.3 


1  - 
0.5- 
0 


- 

/  ^  ^V-:- Cur-Sun 

2yr-Sun 

\ 

-  2yr-Shade 

■N   0 

- 

Cur-Shade 
—  '          '         '  ' 

M  J 

J       A       S  0 

2yr-Shade 


2yr-Sun 


M 


0 


1988 


Figure  5 — Seasonal  changes  in  net  photosyn- 
thetic  rates  {P^J  at  1 ,000  iimol/m^/s  and 
10  °C,  and  dark  respiration  rates  (fl)  at  10  °C 
of  P.  pumila  needles  during  1988.  □  =  current, 
sun-needles;  ■  =  current,  shade-needles; 
O  =  2-year-old,  sun-needles;  •  =  2-year-old, 
shade-needles. 


Needle  longevity  of  P.  pumila,  4  years  or  more,  corre- 
sponds to  that  of  P.  mugo  and  P.  contorta  (Benecke  and 
Havranek  1980). 

Climate  Factors  Affecting 
Photosynthetic  Production 

Seasonal  differences  in  photosynthetic  activity  between 
new  and  old  needles  have  been  reported  for  some  north- 
em  or  subalpine  evergreen  conifers  (Fry  and  Phillips 
1977;  Teskey  and  others  1984).  Likewise,  photos5Tithetic 
activity  (P^^J  in  the  2-year-old  needles  of  P.  pumila  be- 
came higher  between  late  July  and  early  August,  while 
that  of  current  needles  did  so  in  late  August  and  mid- 
September  (fig.  5).  A  similar  seasonal  trend  in  P^^  was 
found  for  the  current  and  older  needles  (1  to  4  years  old) 
of  P.  pumila  in  the  study  plot  in  1987  (Kajimoto  1990). 
These  results  suggest  that  photosynthetic  production  of 
the  P.  pumila  canopy  is  primarily  active  during  August 
and  September. 

Alpine  conifer  trees  rarely  achieve  their  potential  maxi- 
mum photosynthetic  activities  under  natural  environmen- 
tal conditions,  although  the  main  limiting  factors  in  photo- 
sythesis,  such  as  irradiance  and  air  and  soil  temperatures, 
seem  to  be  different  depending  on  species  and  site  condi- 
tions (De  Lucia  and  Smith  1987;  Hasler  1982;  Tiirner  and 
others  1983).  Mean  air  temperatures  between  late  July 
and  mid-September  at  the  research  plot  (fig.  1)  corre- 
sponded to  optimiun  air  temperatm-e  (10-15  °C)  for  net 
photosynthesis  of  P.  pumila,  both  for  current  and  old 


DISCUSSION 

Growth  Characteristics  of  Needles 

Needles  of  P.  pumila  began  to  elongate  in  early  July 
and  almost  finished  growth  in  late  August  (fig.  2).  The 
period  of  needle  growth  is  1  or  2  months  later  than  those 
of  other  low-altitude  pines,  for  example,  P.  densiflora 
(Tanaka  and  others  1976)  and  P.  thunbergii  (Nagatsu 
1987),  growing  naturally  in  central  Japan.  After  needle 
growth  had  terminated,  the  current  needles  of  P.  pumila 
showed  gradual  decline  in  specific  leaf  area  (SLA)  (fig.  3) 
and  maintained  higher  dark  respiration  rates  than 
2-year-old  needles  (fig.  5).  This  indicates  that  the  current 
needles  continued  to  accumulate  internal  carbohydrate  or 
nutrient  until  October. 

In  the  study  area,  there  was  little  local  variation  in  sea- 
sonal growth  patterns  in  the  shoots  and  needles  of  P. 
pumila  (Kajimoto  1989b).  However,  the  final  lengths  of 
current  shoots  and  needles  became  smaller  with  increas- 
ing altitude,  as  reported  for  other  high-altitude  pines, 
such  as  P.  cembra  (Baig  and  Tranquillini  1976),  and  P. 
mugo  and  P.  contorta  (Benecke  and  Havranek  1980). 

The  current  needles  of  P.  pumila  in  the  research  plot 
accounted  for  9  percent  of  total  foliage  biomass  in  late 
July  and  27  percent  in  late  August  (Kajimoto  1989a). 
This  indicates  that  70  percent  or  more  of  the  pine  canopy 
consisted  of  older  needles  throughout  the  growing  season. 


^  August 

^  30- 


u 
c 

Of 
D 
CT 
<1> 


> 

"a 
cr 


20 
10 


^  '0 


10 


September 

30 
20 
10 
0 


20 


Air  temperature  CO 


Figure  6 — Relative  frequency  distributions  of 
daytime  (8  a.m.  to  4  p.m.)  mean  air  tempera- 
ture at  each  1-hour  interval  in  August  (above) 
and  September  (below)  1 988.  Air  tempera- 
tures were  recorded  at  the  P.  pumila  stand 
(2,600  m  a.s.l.),  Kiso  Mountain  Range. 


96 


30- 


o 
c 

u 
cr 


20  - 
10  - 


August 


1000 


2000 


> 


40 


5  301- 
(T  20h 


September 


10  ^ 


0  *^^^^°"'f" 


0  1000  2000 

PFD      ( jjnnol  m'h''  ) 

Figure  7 — Relative  frequency  distributions  of 
daytime  (8  a.m.  to  4  p.m.)  photosyntiietic  photon 
flux  density  (PFD)  at  each  1  -hour  interval  in 
August  (above)  and  September  (below)  1 988. 
PFD  was  measured  above  the  canopy  surface  of 
P.  pumila  stand  (2,600  m  a.s.l.),  Kiso  Mountain 
Range. 

needles  (Kajimoto  1990).  Figure  6  shows  patterns  of  rela- 
tive frequency  distribution  of  daytime  mean  air  tempera- 
ture (8  a.m.  to  4  p.m.)  at  1-hour  intervals;  temperature 
values  between  10  and  16  °C  made  up  about  70  percent  of 
the  sum  total  in  both  August  and  September.  With  re- 
gard to  irradiance  conditions,  proportions  of  PFD  values 
above  600  |j.mol/m7s,  when  the  pine  needles  were  photo- 
synthetically  light  saturated  (table  1),  were  53  percent  in 
August  and  36  percent  in  September  (fig.  7).  The  tem- 
perature and  light  regimes  indicate  that  net  assimilation 
of  P.  pumila  needles  under  natural  conditions  is  limited 
mainly  by  lower  levels  of  solar  irradiance.  Reduction  in 
available  PFD  from  August  (53  percent)  to  September 
(36  percent)  is  likely  to  impede  net  photosynthesis  of  the 
current  needles,  since  photosynthetic  activities  (P^^)  of 
the  current  needles  were  the  highest  in  September  (fig.  5). 

It  is  concluded  from  the  present  study  that  P.  pumila 
needles  are  not  very  efficient  in  photosynthesis  diiring  the 
first  growing  season.  The  pine  needles  survive  for  four  or 
more  growing  seasons  and  recover  their  photosynthetic 
abilities  in  each  summer;  P     of  the  oldest  needles 

'  max 

(4  years  old)  was  about  half  of  that  of  1-year-old  needles, 
although  P^^  decreased  with  needle  age  (Kajimoto  1990). 
Consequently,  a  large  proportion  of  annual  net  carbon 
gain  in  the  P.  pumila  tree  is  likely  to  depend  on  net  as- 
similation by  the  old  needles. 

Knowledge  of  physiological  properties  of  P.  pumila, 
especially  water  relations  (Ando  and  Kawasaki  1991),  is 
still  limited  compared  to  other  subalpine  pine  species.  As 
to  seasonal  photosynthetic  production  of  P.  pumila,  more 
information  on  its  photosynthetic  reponses  to  environ- 
mental factors,  particularly  based  on  field  measurement, 
is  needed. 


ACKNOWLEDGMENTS 

I  thank  Dr.  K.  Hozumi  for  his  suggestions  and  en- 
couragement during  this  study,  and  Drs.  T.  Koike, 
M.  Ishizuka,  A.  Osawa,  Y.  Chiba,  and  N.  Kurachi  for  their 
useful  comments  on  this  manuscript.  Comments  by  Drs. 
F.-K.  Holtmeier,  W.  C.  Schmidt,  and  R.  Hasler  are  ac- 
knowledged. Thanks  also  go  to  Professors  T.  Horiuchi 
and  S.  Ito,  Shinshu  University,  and  to  the  staff  of  Shinshu 
University  Experimental  Forest  and  of  Nagoya  University 
Forest  for  their  assistance  during  the  field  siu"vey. 

REFERENCES 

Ando,  H.;  Kawasaki,  K.  1991.  Growth  of  Pinus  pumila  in 
relation  to  the  site  conditions  on  a  slope.  Transactions 
of  the  102d  Meeting,  Japanese  Forestry  Society: 
451-452.  [In  Japanese]. 

Baig,  M.  N.;  Tranquillini,  W.  1976.  Studies  on  upper  tim- 
ber line:  morphology  and  anatomy  of  Norway  spruce 
(Picea  abies)  and  stone  pine  (Pinus  cembra)  needles 
from  various  habitat  conditions.  Canadian  Journal  of 
Botany.  54:  1622-1632. 

Benecke,  V.;  Havranek,  W.  M.  1980.  Phenological  growth 
characteristics  of  trees  with  increasing  altitude, 
Craigieburn  Range,  New  Zealand.  In:  Benecke,  U.; 
Davis,  M.  R.,  eds.  Mountain  environments  and  subal- 
pine tree  growth.  Tech.  Pap.  70.  New  Zealand  Forest 
Service,  Forest  Research  Institute:  155-174. 

De  Lucia,  E.  H.;  Smith,  W.  K.  1987.  Air  and  soil  tempera- 
ture limitations  on  photosynthesis  in  Engelmann 
spruce  during  summer.  Canadian  Journal  of  Forest 
Research.  17:  527-533. 

Fry,  D.  J.;  Phillips,  I.  D.  J.  1977.  Photosynthesis  of  coni- 
fers in  relation  to  annual  growth  cycles  and  dry  matter 
production.  II.  Seasonal  photosynthetic  capacity  and 
mesophyll  ultrastructure  in  Abies  grandis,  Picea 
sitchensis,  Tsuga  heterophylla  and  Larix  leptolepis  in 
S.W.  England.  Physiologia  Plantarum.  40:  300-306. 

Hadley,  J.  L.;  Smith,  W.  K.  1987.  Influence  of  krummholz 
mat  microclimate  on  needle  physiology  and  survival. 
Oecologia.  73:  82-90. 

Hasler,  R.  1982.  Net  photosjTithesis  and  transpiration  of 
Pinus  montana  on  east  and  north  facing  slopes  at  al- 
pine timberline.  Oecologia.  54:  14-22. 

Holtmeier,  F.-K.  1973.  Geoecological  aspects  of  timber- 
lines  in  northern  and  central  Europe.  Arctic  and  Alpine 
Research.  5(3):  45-54. 

Holtmeier,  F.-K.  1981.  What  does  the  term  "krummholz" 
really  mean?  Observations  with  special  reference  to 
the  Alps  and  the  Colorado  Front  Range.  Mountain 
Research  and  Development.  1(3/4):  253-260. 

Kajimoto,  T.  1989a.  Aboveground  biomass  and  litterfall  of 
Pinus  pumila  scrubs  growing  on  the  Kiso  Mountain 
Range  in  central  Japan.  Ecological  Research.  4:  55-69. 

Kajimoto,  T.  1989b.  Current  growth  of  Pinus  pumila. 
Transactions  of  the  37th  Meeting,  Chubu  Branch, 
Japanese  Forestry  Society:  55-58.  [In  Japanese]. 

Kajimoto,  T.  1990.  Photosynthesis  and  respiration  of 
Pinus  pumila  needles  in  relation  to  needle  age  and 
season.  Ecological  Research.  5:  333-340. 


97 


Kajimoto,  T.  1992.  Dynamics  and  dry  matter  production 
of  belowground  woody  organs  of  Pinus  pumila  trees 
growing  on  the  Kiso  Mountain  Range  in  central  Japan. 
Ecological  Research.  7:  333-339. 

Larcher,  W.  1975.  Physiological  plant  ecology.  Heidelberg: 
Springer.  252  p. 

Nagano  Meteorological  Observatory.  1988.  Annual  report 
on  weather  conditions  of  Nagano  prefecture  in  1987. 
Nagano  Branch  Office  of  Japanese  Meteorological  Soci- 
ety. 137  p.  [In  Japanese]. 

Nagatsu,  M.  1987.  Correlation  between  the  growth  of 
needles  and  shoots  in  a  Pinus  thunbergii  yoimg  tree. 
Transactions  of  the  98th  Meeting  Japanese  Forestry 
Society:  417-418.  [In  Japanese]. 

Okitsu,  S.;  Ito,  K.  1984.  Vegetation  dynamics  of  the  Sibe- 
rian dwarf  pine  {Pinus  pumila  Regel)  in  the  Taisetsu 
Mountain  Range,  Hokkaido,  Japan.  Vegetatio.  58: 
105-113. 

Okitsu,  S.;  Ito,  K.  1989.  Conditions  for  the  development  of 

the  Pinus  pumila  zone  of  Hokkaido,  northern  Japan. 

Vegetatio.  84:  127-132. 
Sakai,  A.;  Larcher,  W.  1987.  Frost  survival  of  plants.  Ecol. 

Stud.  62.  New  York:  Springer.  321  p. 
Shidei,  T.  1963.  Productivity  of  Haimatsu  {Pinus  pumila) 

community  growing  in  alpine  zone  of  Tateyam-Range. 


Journal  of  the  Japanese  Forestry  Society.  45:  169-173. 
[In  Japanese  with  English  simimary]. 

Tanaka,  H.;  Oohata,  S.;  Akai,  T.  1976.  Elongation  and 
shoot  forms  in  foreign  pines.  Report,  Kyoto  University 
of  Forestry.  11:  38-49.  [In  Japanese]. 

Teskey,  R.  O.;  Grier,  C.  C;  Hinckley,  T.  M.  1984.  Changes 
in  photosynthesis  and  water  relations  with  age  and  sea- 
son in  Abies  amabilis.  Canadian  Journal  of  Forest 
Research.  14:  77-84. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Ecol.  Stud.  31.  New  York:  Springer.  131  p. 

Turner,  H.;  Hasler,  R.;  Schonenberger,  W.  1983.  Contrast- 
ing microenvironments  and  their  effects  on  carbon  up- 
take and  allocation  by  young  conifers  near  alpine 
treeline  in  Switzerland.  In:  Wareing,  R.  H.,  ed.  Carbon 
uptake  and  allocation  in  subalpine  ecosystems  as  a  key 
to  management:  Proceedings  of  lUFRO  workshop.  1982 
August  2-3.  Corvallis,  OR:  Oregon  State  University, 
Forest  Research  Laboratory:  22-30. 

Wardle,  P.  1977.  Japanese  timberlines  and  some  geo- 
graphic comparisons.  Arctic  and  Alpine  Research.  9: 
249-258. 


98 


HEIGHT  GROWTH  IN  CEMBRAN  PINE 
AS  A  FACTOR  OF  AIR  TEMPERATURE 

Herbert  Kronfuss 


Abstract — The  study  treats  height  growth  in  cembran  pine 
iPinus  cembra)  in  high-elevation  afforestation  at  an  altitude 
of  1,800  m  with  growth  patterns  analyzed  as  a  factor  of  air  tem- 
perature. It  is  assumed  that  growth  behavior  follows  an  endog- 
enous rhythm  possibly  deriving  from  the  genotj^je.  The  relation- 
ship between  growth  pattern  and  air  temperature  is  demon- 
strated for  defined  periods  and  phases  of  height  growth.  From 
the  results  it  becomes  obvious  that  height  growth  and  thus  the 
relative  rate  of  growth  are  primarily  dependent  on  temperature. 
Data  from  long-term  monitoring  series  were  used  to  determine 
the  ampUtudes  between  the  growth  curves  for  early  and  for  late 
culmination.  A  comparison  of  the  ampUtudes  for  cembran  pine 
with  spruce  and  larch  shows  cembran  pine — with  the  highest 
amplitudes — to  be  the  most  sensitive  to  temperature.  Cembran 
pine  thus  makes  most  efficient  use  of  warm  weather  periods  for 
increment  production  in  short  periods  of  time. 


The  cembran  pine  (Pinus  cembra)  occupies  the  highest 
stands  of  all  trees  in  the  Alps  and  as  such  is  ideally  suited 
for  a  study  of  height  growth  related  to  air  temperature. 
The  fact  that  cembran  pine  is  not  thought  to  have  differ- 
ent ecotypes  adapted  to  various  altitudes  also  permits 
growth  characteristics  of  this  species  to  be  defined  in  spe- 
cific terms. 

STUDY  SITE 

The  study  site  was  located  30  km  southwest  of  Innsbruck, 
Tyrol  (Austria),  in  the  Stubai  Alps  near  Haggen  in  the 
Sellrain  Valley.  For  location,  geology,  and  climate  details, 
see  figures  1  and  2,  and  table  1.  The  results  relate  to  a 
continuous  experimental  stand  of  cembran  pine  at  an  alti- 
tude of  1,800  m.  The  stand  was  established  in  1970  when 
the  plot  was  afforested  with  4-year-old  plants. 

MONITORING 

Monitoring  was  performed  over  a  period  of  8  years 
for  a  group  of  40  trees.  Height  growth  in  cembran  pine 
was  measured  at  5-day  intervals  (pentades)  and  the  rate 
of  growth  correlated  with  mean  air  temperature  of  the 
pentade.  Individual  height  increments  were  expressed 
as  a  percentage  of  annual  cumvdative  grov^^h  to  permit 
the  growth  patterns  to  be  compared  -with,  changes  in 
temperatxire. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Herbert  Kronfuss  is  Forester,  Forestry  Research  Institute  of  Austria, 
Rennweg  1/29,  A-6020  Innsbruck. 


Growth  patterns  are  represented  in  the  form  of  relative 
growth-rate  c;irves  and  relative  cumulative-growth  curves. 
This  not  only  permits  direct  comparisons  to  be  made  be- 
tween annual  increments  but  also,  given  a  long  enough 
monitoring  period,  permits  a  characteristic  increment  pat- 
tern (type)  to  be  derived  for  each  species  (Burger  1926; 
Larcher  1980;  Week  1955). 

HEIGHT  GROWTH 

From  1978  to  1985  increment  growth  lasted  an  average 
of  108  days,  vnth  a  20-day  variance  range. 

To  study  the  influence  of  temperature  on  the  growth 
pattern  during  the  period  of  extension  growth,  the  rela- 
tive grovrth-rate  curve  was  correlated  with  the  tempera- 
ture curve,  with  the  mean  height  increments  from  con- 
secutive pentades  expressed  as  a  percentage  of  total 
height  growth.  The  relative  growth-rate  curve  clearly 
illustrates  the  pattern  of  height  grov^h  (fig.  3). 

Figure  3  shows  how  temperature  fluctuations  during 
the  period  of  extension  growth  influence  the  rate  of  height 
growth.  With  regard  to  the  correlation  between  the 
height-growth  behavior  and  temperattire  patterns,  the 
following  points  would  seem  to  be  of  interest. 

Dming  growing  season,  the  temperature  curve  itself 
is  characterized  by  two  pronotmced  dips  for  the  periods 
May  20  to  25  and  Jtme  15  to  20,  a  phenomenon  well 
knovm  to  the  local  rural  communities  as  the  "Ice  Saints" 
and  "Sheep's  Chill."  At  the  beginning  of  April,  mean  tem- 
perattire  reaches  freezing  point  on  the  long-term  average. 
The  rise  in  temperature  continues  imtil  mid-May,  reach- 
ing an  average  of  approximately  7  °C,  before  dropping 
back  to  5  °C  for  the  pentade  occupying  the  first  of  the  two 


Figure  1 — Geographical  location  of  the  site. 


99 


HAGGEN  i.S.dSOO  m)  a)  3,1°  b)  920  mm 
c)  10         d)  1975  -  1984 


mm 
200 

100 


10 


I  I  I  I  I 


I  I  I  I  I  I  I 


I  I  I  I  I 


M 


M 


a)  Mean  annual  temperature 

b)  Mean  annual  precipitation 

c)  Number  of  years  of  data  collection  for  temperature  and 
precipitation 

d)  Period  of  data  collection  for  temperature  and 
precipitation:  1975-1984 

e)  Mean  monthly  precipitation 

f)  Mean  monthly  temperature 

g)  Months  with  average  daily  minimum  below  0°C 

h)  Months  with  absolute  minimum  temperatures  below  0°C 

Figure  2 — Data  for  Haggen  Climate  Diagram 
Haggen  I.S.,  Field  Station  1, 1,800  m,  SSW. 


cold  spells.  During  the  time  of  the  "Ice  Saints,"  most  of 
the  cembran  pines  are  at  the  beginning  of  the  period  of  ex- 
tension growth,  with  a  growth  rate  of  only  2  to  3  percent 
of  relative  growth  per  pentade.  That  corresponds  to  about 
1.5  cm.  On  the  average,  approximately  7  percent  of  cumu- 
lative growth  is  achieved  by  this  point  in  time.  For  height 
grovi^h  in  cembran  pines  in  this  location,  the  period  be- 
tween the  "Ice  Saints"  and  "Sheep's  Chill"  is  of  special 
importance  as  it  includes  the  first  major  peak  in  the  tem- 
perature curve,  on  or  around  May  31  on  average. 

With  a  brief  delay,  this  temperature  peak  stimulates 
a  response  in  the  cembran  pine  in  terms  of  height  growth, 
and  after  one-  and  one-half  pentades,  the  biggest  absolute 
increment  is  produced,  accounting  for  about  35  percent  of 


total  growth  on  a  long-term  average.  Following  this  peak 
in  increment  production,  the  typical  pattern  of  height 
growth  in  the  cembran  pine  is  one  of  declining  rates  of 
growth  in  spite  of  rising  temperatures.  This  is  shown  in 
figure  3  with  the  two  curves  moving  in  opposite  directions 
following  the  start  of  the  "Sheep's  Chill"  period  aroimd 
June  20. 

Not  even  an  increase  in  mean  pentade  temperature 
beyond  the  10  °C  mark  around  July  5,  finally  peaking 
at  13  °C,  leads  to  a  further  spontaneous  spate  of  growth. 


Table  1 — Location  of  the  study:  HAGGEN  afforestation  site  in  the 
Sellrain  Valley  near  St.  Sigmund,  Tyrol 


Geographical  coordinates: 
Height  above  sea-level: 
Exposition  and  angle  of  slope: 


47°13'N.  ir06'E. 
1,715-1,950  m 
SSW,  30° 


Bedrock:         Foliated  gneiss  (biotite  granite  gneiss,  granodiorite 
gneiss)  and  mica  schist. 

Vegetation  at 

afforestation:  Heath  (callunetum,  nardetum  alpigenum) 
Soils:  Humus  horizons  developed  into  a  brown  earth  type 

through  grazing  and  haymaking.  More  or  less 

podsolic  brown  soils. 
In  previous  centuries  a  heavily  grazed  avalanche  slope. 

CLIMATIC  DATA  are  based  on  a  10-year  monitoring  period  (1975- 
1984) 

-  Sunshine  hours  during  the  vegetation  period  (mid-May  to 

September) 

a)  theoretical  asatronomical  maximum  2,005  hours 

b)  theoretical  local  maximum  1 ,393  hours 
horizon  loss                                         30.5  percent 

-  Air  temperature 

a)  mean  annual  temperature  3.1  °C 

b)  during  vegetation  period  8.4  °C 

-  Precipitation 

a)  mean  annual  precipitation  (3  m  aboveground)         920  mm 

b)  during  vegetation  period  (3  m  aboveground)  533  mm 

c)  precipitation  at  ground  level  650  mm 
(=  approx.  22  percent  higher  than  at 

3  m  aboveground) 

-  Precipitation  probability  (percent) 

a)  June  maximum  55  percent 

b)  December  minimum  27  percent 

-  Mean  precipitation  distribution  (mm) 

a)  maximum  in  July  and  September  8.4  mm 

b)  minimum  in  February  3.6  mm 

c)  month  with  highest  precipitation  (July)  141  mm 

d)  month  with  lowest  precipitation  (February)  33  mm 

-  Evaporation  during  the  vegetation  period 

a)  potential  evaporation  (2.5  m  aboveground  566  mm 
with  "piche")  (=  4.1  l/m^  per  day) 

b)  actual  evapotranspiration  (lysimeter  with  236  mm 
ground  vegetation  (=  1 .7  l/m^  per  day) 

c)  actual  hydrological  budget  412  mm 
(=  3  l/m^  per  day) 

The  meteorological  summer  semester  accounts  for  65  percent  of 
annual  precipitation  (598  mm),  and  the  meteorological  winter 
semester  35  percent  (322  mm). 


100 


Mean  pentad  values  for  relative  growth  rate 
and  air  temperature  (°C) 


Mean  relative  growth  rate  (%) 


January  February    March     April        May      June       July     August  September  October  November  December 

Figure  3 — Comparison  between  the  curves  of  relative  growth  rate  and  air 
temperature,  8-year  series  (1978  to  1985). 


The  minor  fluctuations  in  the  growth  curve  on  the  order 
of  1  to  3  percent  can  be  correlated  with  temperature  fluc- 
tuations of  10  to  13  °C. 

The  main  features  of  the  curves  in  figure  3  can  be  de- 
scribed as  follows:  The  grovvi;h  curve  for  cembran  pine 
peaks  (maximum  growth  rate  per  pentade  is  reached) 
on  June  5  to  10  on  average.  The  temperature  curves  are 
characterized  by  two  peaks,  the  first  between  the  two  pro- 
nounced dips  (May  31  to  June  5)  and  the  second  between 
July  25  and  July  31.  It  is  interesting  to  note  that  height 
increments  between  July  5  and  the  end  of  the  period  of 
extension  growth — and  that  is  a  good  46  days — total 
only  20  percent  of  cumulative  grov\i;h,  even  though  the 
temperature  curve  passes  10  °C  on  July  5  and  remains 
more  or  less  at  that  level  until  mid-September. 

The  above  picture,  based  on  the  mean  figures  calculated 
from  monitored  data,  illustrates  in  bold  strokes  the  corre- 
lation between  the  temperature  and  height-growth  patterns. 

The  relative  grov^i:h-rate  curves  shown  in  figure  4  for 
the  8  years  from  1978  to  1985  provide  an  overview  of 
growth  patterns  in  cembran  pine  as  a  factor  of  air  tem- 
perature for  the  individual  years.  One  of  the  more  strik- 
ing features  of  these  curves  is  the  variation  in  the  onset 
of  culmination  (equals  greatest  increment  production  per 
pentade).  On  this  basis  we  can  distinguish  between  an- 
nual growth  curves  with  early  culmination  (1979,  1981, 
1982,  1983),  one  year  with  late  culmination  (1984),  and 
years  with  twin  or  intermediate  peaks  (1978,  1980,  1985; 
see  also  table  2). 

A  comparison  of  the  temperature  curve  in  figure  3  with. 
the  growth-rate  curves  in  figure  4  reveals  two  clearly 
delineated  time  periods  in  which  culmination  (peak 
growth  rate  per  pentade)  occurs.  The  first  is  the  period 
between  the  "Ice  Saints"  and  "Sheep's  Chill,"  and  the  sec- 
ond is  a  subsequent  period  of  30  days  ending  approxi- 
mately with  the  pentade  of  July  15  to  20.  These  two  peri- 
ods together  total  57  days  (equals  53  percent  of  the  mean 


 1978 

...z 

■  1  

. . . .  1  

!^^^7^  ^ 

10  20  31   10  20  30  10  20  31  10  20  31 

May  June      July  August 


W  20  31  10  20  30  10  20  31  10  20  31 


—  »« 

J. 

■   

^^~.yrr>^  

10  20  31   10  20  30  10  20  31  10  20  31 


 1980 

■  ■1  



W  20  31   10  20  30  10  20  31  10  20  31 


 1981 

■A^  

«  20  31   10  20  30  10  20  31  10  20  31 


 1983 

,,l  



10  20  31   10  20  30  10  20  31   10  20  31 


10  20  31   10  20  30  10  20  31   10  20  31 


 1986 

10  20  31  10  20  30  10  20  31  10  20  31 


Figure  4 — Cembran  pine  relative  height-growth 
curves  (percent). 


101 


Table  2 — Air  temperature  in  °C  and  percentage  growth  in  cembran 
pine' 


J   n  I  I 


Year 

TS 

T 

TS 

T 

HG 

HG 

 °C- 

-  -  Percent  -  - 

Early  culmination 

1979 

245 

9.1 

298 

9.9 

71 

22 

1981 

242 

8.9 

230 

7.7 

61 

19 

1982 

246 

9.1 

368 

12.3 

52 

O  IT 

35 

1983 

203 

7.5 

359 

12.0 

46 

38 

Average 

234 

8.7 

314 

10.5 

58 

28 

Intermediate  culmination 

8.2 

241 

8.0 

38 

52 

1980 

195 

7.2 

205 

6.8 

26 

47 

1985 

191 

7.1 

284 

9.5 

31 

50 

Average 

203 

7.5 

243 

8.1 

32 

50 

Late  culmination 

1984 

169 

6.3 

284 

9.5 

17 

58 

'I  =  period  from  the  "Ice  Saints"  to  "Sheep's  Chill";  II  =  period  from  "Sheep's 
Chill"  to  20  July;  TS  =  temperature  sum;  T  =  mean  temperature;  HG  =  height 
growth;  period  I  lasts  for  27  days;  period  II  lasts  for  30  days. 


10      20         31         10         20         30         10         20        31         10        20  31 

May  June  July  August 


Figure  5 — Cembran  pine  relative  height  growth 
(percent). 


period  of  extension  growth)  and  account  for  an  average 
of  82  percent  of  annual  shoot  length. 

There  is  a  significant  correlation  between  percentage 
growth  within  the  two  main  growth  periods  and  tempera- 
ture sum,  v^dth  peak  growth  located  in  the  period  that  ex- 
hibits the  highest  temperature  sum  in  absolute  terms, 


except  that  beyond  an  average  temperature  sum  of  230  °C 
for  the  first  period  a  further  increase  in  temperature  for 
the  second  period  does  not  trigger  higher  growth  rates  in 
that  period.  The  highest  pentade  growth  rate  always  oc- 
curs in  the  period  with  the  highest  temperature  Bum  and 
coincides  with  or  follows  the  pentade  with  maximum 
pentade  temperature  (see  table  2). 

From  table  2  it  can  be  seen  that  in  the  case  of  early  cul- 
mination the  average  temperature  sum  in  the  first  period 
was  234  °C  (8.7  °C/27  days).  During  that  period,  an  aver- 
age of  58  percent  of  extension  growth  was  measured  (with 
scatter  between  46  percent  and  71  percent).  In  those 
cases  the  cembran  pines  achieved  an  average  of  only  28 
percent  relative  growth  in  the  second  period  in  spite  of  the 
rise  in  the  average  temperature  sum  to  314  °C  (10.5  °C/30 
days). 

The  year  of  late  culmination,  1984,  is  also  instructive. 
The  cold  weather  in  period  I  of  that  year  produced  a  tem- 
perature sum  of  only  169  °C  (6.3  °C/27  days),  which  is 
reflected  in  a  correspondingly  low  relative  growth  rate 
of  17  percent.  Period  II,  however,  was  characterized 
by  a  significant  rise  in  temperature,  producing  a  tem- 
perature sum  of  284  °C  (9.5  °C/30  days).  The  relative 
growth  rate  in  that  period  (fig.  5)  was  58  percent,  which 
is  identical  with  the  figure  for  period  I  in  the  years  of 
early  culmination. 

At  the  end  of  July  in  the  late  culmination  year, 
the  leaders  had  reached  90  percent  of  total  extension 
growth  compEired  with  98  percent  for  the  years  of  early 
culmination. 

This  analysis  of  the  height-growth  patterns  of  cembran 
pine  over  an  8-year  period  of  observation  (1978  to  1985) 
shows  that  the  cembran  pine  is  ontogenetically  equipped 
to  make  efficient  use  of  available  warmth  for  increment 
production  so  as  to  achieve  ctdmination  as  early  as  pos- 
sible. Up  to  that  point,  air  temperature  is  the  dominant 
factor  in  determining  the  relative  growth  rate,  £ind  after- 
ward it  continues  to  function  as  the  "engine"  that  main- 
tains the  extension  process. 

PHASES  OF  HEIGHT  GROWTH 

The  following  analysis  of  annual  height  growth  by  help 
of  the  growth  curve  (Kronfuss  1985)  will  illustrate  the 
somewhat  abstract  correlation  between  growth  rate  and 
temperature  with  the  aid  of  further  figures.  For  this  pur- 
pose, height  growth  can  be  divided  into  the  following 
three  phases  (fig.  6): 

a.  Initial  phase  (0  to  25  percent  of  annual  extension). 

b.  Main  phase  (25  to  75  percent  of  annued  extension). 

c.  Final  phase  (75  to  100  percent  of  annual  extension). 

Table  3  provides  a  separate  assessment  of  the  correla- 
tion for  each  of  these  three  phases  using  a  regression  line, 
correlation  coefficient,  and  regression  coefficient  to  ex- 
press the  relationship  between  mean  daily  temperature 
(T/d)  and  average  height  growth  per  day  (HG/d)  (fig.  6, 
table  3). 

Whereas  the  correlation  coefficient  for  the  initial  phase 
(r  =  0.6)  suggests  a  correlation  vnth  temperature,  this  is 
no  longer  the  case  in  the  final  phase  (r  =  0.42).  The  main 


102 


Correlation  between  mean  daily  air  temperature  and  height  increment 
per  day  HAGGEN  i.S.,  1800  m 

Average  height  increment  per  day  (mm) 


4- 

r  =  0.89  k=^^!£A 

^      Mam  phase 

Initial  phase 

r=.0.42  /f.O.II 

• 

•  • 

r=o.60  ;f=o.is  ■ 

•  •• 

 1  i_ 

*           "  *  Final  phase 

_      1  ■ 

S           6             7  8 

9  10 

It             12           13  14 

Air  temperature  ("C) 


Correlation  between  mean  daily  air  temperature  and  growth  processes 
HAGGEN  i.S..  1800  m 

Duration  of  phases  (days) 


Air  temperature  (°C) 

Figure  6 — Cembran  pine  correlations  between 
mean  daily  air  temperatures,  height  increment,  and 
duration  of  growth  phases:  period  1 978  to  1 985. 


growth  phase,  on  the  other  hand,  is  largely  determined  by 
temperature,  as  can  be  seen  from  the  almost  stretched 
curve  (fig.  6  at  top)  and  a  correlation  coefficient  of  r  =  0.89. 

The  regression  coefficients  {h),  as  the  expression  of  aver- 
age daily  growth  per  vinit  of  temperature,  show  that,  for 
the  same  temperature  rise,  height  growth  in  the  main 


phase  is  4.5  times  greater  than  in  the  initial  phase  and 
6.2  times  greater  than  in  the  final  phase.  In  the  initial 
phase,  with  an  average  period  of  extension  growth  of  32 
days,  the  average  daily  increment  is  1.6  mm.  In  the 
main  growth  phase  the  period  of  extension  growth  is  re- 
duced to  an  average  of  26  days  of  an  HG/d  of  4  mm,  while 
the  figure  for  the  subsequent  50  days  of  the  final  phase  is 
only  1  mm. 

The  duration  of  extension  growth  also  correlates  with 
temperature  and  is  shorter  with  higher  temperature  rise. 
The  influence  of  temperature  on  the  period  of  extension 
growth  can  be  seen  from  the  correlation  coefficients,  with 
a  close  correlation  with  temperature  expressed  by  the  co- 
efficients r  =  0.82  for  the  initial  phase  and  r  =  0.84  for 
the  main  growth  period,  decreasing  in  the  final  phase 
to  r  =  0.59.  This  low  correlation  coefficient  shows  that 
reduced  height  growth  in  the  final  phase  is  no  longer  due 
exclusively  to  the  influence  of  temperature,  but  to  the 
tailing  away  of  an  endogenous  pattern.  With  regard  to 
the  temperature-dominated  phases  of  extension  growth, 
(the  initial  and  main  phases)  an  increase  in  mean  tem- 
perature by  1  °C  reduces  the  extension  growth  period  by 
4  to  5  days.  Also,  in  the  final  phase  we  see  again  that  the 
pattern  derives  from  an  autonomous  endogenous  rhythm 
that  determines  the  potential  period  of  extension  growth 
in  this  phase  for  the  species  involved. 

GROWTH  CURVE  VARIANTS 

As  can  be  seen  from  the  shapes  of  the  relative  growth- 
rate  curves  in  figure  4,  annual  patterns  of  height  growth 
vary  considerably  for  the  same  species  in  the  same  loca- 
tions. Eight  years  of  observation  offer  enough  data  for 
graphic  representation  of  maximum  amplitudes  between 
the  growth  curves  for  the  earliest  and  those  for  the  latest 
culminations  and  for  a  comparison  with  spruce  and  larch 
(fig.  7).  This  illustrates  not  only  the  time  sequences  with 
regard  to  flushing  but  also  the  cxirrent  range  of  fluctua- 
tion for  these  tree  species. 


Table  3 — Correlations  between  mean  daily  temperature  (t7d)  and  height  growth  per  day  (HG/d)  and  between  mean  dally 
temperature  (t7d)  and  the  period  of  extension  growth  days  (d) 


Year 

Initial  phase^ 

Main  phase^ 

Final  phase^ 

t°/d 

HG/d 

d 

t°/d 

HG/d 

d 

t°/d 

HG/d 

d 

1978 

7.9 

1.8 

24 

6.9 

2.6 

33 

10.4 

0.9 

48 

1979 

9.6 

1.8 

22 

8.8 

4.7 

17 

9.9 

.6 

71 

1989 

5.4 

.9 

43 

6.6 

2.5 

32 

12.0 

1.0 

42 

1981 

6.5 

1.7 

25 

.4 

4.7 

18 

8.9 

.7 

59 

1982 

8.2 

1.6 

28 

9.3 

4.1 

22 

11.7 

.9 

52 

1983 

6.1 

1.8 

33 

8.8 

4.6 

26 

13.1 

1.5 

40 

1984 

5.2 

1.3 

52 

9.7 

5.5 

25 

9.8 

1.6 

42 

1985 

7.5 

2.1 

29 

7.2 

3.5 

35 

11.2 

1.4 

43 

Average 

7.1 

1.6 

32 

8.5 

4.0 

26 

10.9 

1.1 

50 

'Initial  phase:  correlation  coefficient  r=  0.604;  regression  coefficient  /( =  0.147;  point  of  intersection  b=  0.587;  standard  deviation  s=  0.326. 
^Main  phase:  r  =  0.889;  k  =  0.681 ;  b  =  -1 .471 ;  s  =  1 .000. 
^Final  phase:  r  =  0.426;  /c  =  0. 1 1 7;  b  =  -0. 1 95;  s  =  0.353. 


103 


Cumulative  growth  (%)  / 

 ^  -   -.--^ 

—  ^  V/- 

/   0  / 
  /    cr    y  / 

/-/"  /•  

/ '''' 

tj~r" 

r  1  1  1 

1     1     L     1     1     1     J     1     1     1     J     1     1  1 

1     10    20   30    10    20    31    10    20    30    10    20    31    10    20    31    10    20  30 
April  Sept. 

Figure  7 — Characteristic  growth  curve  ampli- 
tudes for  cembran  pine,  spruce,  and  European 
larch:  afforestation  "Haggen"  in  1 ,800  m  a.s.l. 


Apart  from  the  differences  in  the  beginning  and  end  of 
the  period  of  extension  growth,  cembran  pine  is  the  most 
sensitive  to  temperature  of  the  three  species,  as  the  high 
amplitude  shows.  It  can  be  concluded  that  the  cembran 
pine  makes  the  most  efficient  use  of  warm  weather  condi- 
tions for  increment  production.  In  the  initial  phase  of 


extension  growth,  cembran  pine  is  already  well  ahead  of 
the  other  species  in  terms  of  growth.  In  the  main  growth 
phase,  larch  is  also  surpassed  by  the  cembran  pine,  while 
spruce  comes  very  close  to  it. 

For  the  main  growth  period,  the  two  extremes  of  the 
growth  curves  for  spruce  and  larch  are  almost  parallel. 
This  shows  that  increment  production  per  pentade  is 
more  balanced  between  warm  and  cold  conditions  than 
in  the  case  of  cembran  pine,  which  is  much  more  sensitive 
to  temperature  fluctuations. 

REFERENCES 

Burger,  H.  1926.  Untersuchungen  iiber  das  Hohen- 
wachstum  verschiedener  Holzarten.  Mitt.  Schweiz. 
Centralanst.  Forstl.  Vers,  wesen;  XIV/I. 

Kronfuss,  H.  1985.  Die  Zuwachsleistimg  einer  Hochlagen- 
aufforstung  mit  Zirbe  (Pinus  cembra)  auf  einem 
Siidhang,  in  Abhaangigkeit  von  der  Seehohe.  lUFRO 
Meeting  Proceedings.  Davos,  Switzerland. 

Larcher,  W.  1980.  Okologie  der  Pflanzen  auf 
physiologischer  Grundlage.  Ulmer,  Stuttgart. 

Week,  J.  1955.  Forstliche  Zuwachs  und  Ertragskimde. 
Neimiann,  Berlin. 


104 


SURVIVAL  AND  GROWTH  OF  PLANTED 
CEMBRAN  PINES  AT  THE  ALPINE 
TIMBERLINE 


Josef  Senn 

Walter  Schonenberger 
Ueli  Wasem 


Abstract — In  an  experimental  afforestation  with  cembran  pines 
{Pinus  cembra  L.)  at  the  subalpine  timberhne,  tree  survival  was 
mainly  determined  by  the  date  of  disappearance  of  the  snow 
cover  in  spring.  Trees  survived  well  on  early  snow-free  sites  and 
poorly  on  sites  with  long-lasting  snow  cover.  The  major  causes  of 
mortality  were  the  two  parasitic  fungi  Ascocalyx  abietina  and 
Phacidium  infestans.  Tree  height,  as  an  index  for  growth,  was 
dependent  on  the  amount  of  energy  available  during  the  vegeta- 
tion season,  and  therefore  negatively  correlated  with  both  alti- 
tude above  sea  level  and  wind  velocity,  and  positively  with  global 
radiation. 


In  the  severe  winter  of  1951-52  avalanches  caused 
major  destruction  all  over  the  Alps.  This  initiated  a 
multidisciplinary  research  program  to  restore  the  upper 
treeline.  The  aim  of  the  program  was  to  develop  methods 
to  alforest  treeless  avalanche  catchments  and  abandoned 
meadows  within  the  subalpine  forest  belt  in  order  to  es- 
tablish new  protection  forests. 

Protection  forests  are  an  important  feature  in  alpine 
landscapes.  Many  areas  can  only  be  permanently  inhab- 
ited as  long  as  these  forests  remain  intact.  In  some  of 
these  forests,  however,  severe  problems  concerning  stand 
stability  have  become  evident.  Inadequate  plantation 
techniques  in  the  past  have  frequently  led  to  forests  that 
were  highly  susceptible  to  natural  hazards  such  as 
storms.  On  the  other  hand,  locally  dense  ungulate  popu- 
lations prevented  successful  natiu-al  rejuvenation  in  the 
existing  forests. 

Both  natural  regeneration  and  artificial  afforestation 
generally  have  become  more  difficult  toward  the  alpine 
timberline,  since  all  biological  processes  become  slower 
with  increasing  altitude  above  sea  level.  In  general,  com- 
pared to  natiiral  invasion  of  forest  trees,  planting  reduces 
the  time  needed  to  establish  forests  that  fulfill  their  pro- 
tective function.  Therefore  artificial  afforestation  may  be 
highly  desirable.  Poor  financial  resources,  however,  may 
Umit  planting  to  the  most  promising  sites. 

The  experimental  afforestation  at  StiUberg  was  designed 
to  obtain  information  about  site  conditions  that  determine 
success  or  failure  of  future  afforestations,  and  to  be  able 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Josef  Senn,  Walter  Schonenberger,  and  Ueli  Wasem  are  members  of  the 
High  Altitude  Afforestation  Group  at  the  Swiss  Federal  Institute  for  For- 
est, Snow  and  Landscape  Research,  CH-8903  Birmensdorf,  Switzerland. 


to  inform  forestry  personnel  which  sites  would  be  most 
promising  for  successful  planting  and  on  which  sites 
planting  would  be  fruitless. 

Cembran  pine  {Pinus  cembra  L.)  is  one  of  the  three  tim- 
berline species  tested  at  StiUberg.  This  species,  extending 
to  the  highest  altitudes  of  any  trees  in  the  Alps,  played  an 
important  role  in  the  culture  of  local  human  populations 
(Holtmeier,  these  proceedings),  produces  valuable  timber, 
and  is  esteemed  by  "green"  tourism  as  an  important  fea- 
ture of  natural  landscapes.  Nonsustainable  exploitation 
in  medieval  times  and  pasturalism  reduced  cembran  pine 
to  a  fraction  of  its  original  abundance,  although  there  is 
evidence  for  natural  range  expansion  in  the  past  100  years. 
Artificial  reproduction  and  estabhshment  of  cembran  pine, 
on  the  other  hand,  pose  many  problems  for  practical  for- 
estry (Frehner  and  Schonenberger,  these  proceedings). 

MATERIAL  AND  METHODS 

The  present  study  was  conducted  at  StiUberg  experi- 
mental area,  a  highly  structiu-ed  northeast  slope  (fig.  1) 
in  the  Dischma  Valley  near  Davos,  Switzerland  (lat. 
34°08'  S.,  long.  171°41'  E.).  The  area  extends  over  9.9  ha 
between  2,080  and  2,230  m  above  sea  level.  The  highest 
parts  reach  beyond  the  altitudinal  timberline.  Before 
plantation  in  1975  various  site  aspects  like  slope,  expo- 
sure, global  radiation  (Turner  1966),  wind  conditions 
(Nageli  1971),  date  of  disappearance  of  snow  cover  in 
spring  (Rychetnik  1987),  ntunber  of  snow-fi-ee  days  in 
winter,  avalanche  fi'equencies,  soil  types  (Blaser  1980), 
and  vegetation  (mainly  dwarf  shrub  communities)  (Kuoch 
1970)  were  recorded  on  a  very  fine  scale.  Schonenberger 
(1975)  described  the  site  in  detail.  Results  from  the 
StiUberg  experimental  afforestation  were  published  by 
Schonenberger  and  others  (1988).  Practical  aspects  for 
high-altitude  forestry  were  included  in  Schonenberger 
and  others  (1990). 

Three  coniferous  tree  species  were  planted  in  1975: 
5-year-old  cembran  pines,  3-year-old  mountain  pines  (P. 
uncinata  (Miller)  Domin  =  Pinus  montana  Miller  =  Pinus 
mugo  Turra  ssp.  uncinata  Domin)  the  erect  form  of  the 
mugo  pine,  and  1-year-old  European  Isirches  (Larix  de- 
cidua  Miller).  Before  planting,  the  whole  area  was  di- 
vided into  square  units  of  3.5  by  3.5  m  (fig.  1).  In  each 
square  unit  25  trees  of  one  species  were  planted.  The 
number  of  trees  per  imit  may  have  been  reduced  by  ad- 
verse local  soil  conditions  like  barren  rock.  The  squares 
contained  alternatively  cembran  pines,  mountain  pines, 
and  larches.  Altogether  some  90,000  trees  were  planted 
in  4,052  square  units. 


105 


Figure  1 — Topographic  model  of  the  Stillberg 
experimental  area  with  the  grid  of  the  affores- 
tation units  (3.5  by  3.5  m). 


This  paper  reports  the  results  of  survival  of  the  cembran 
pines  in  comparison  with  mountain  pine  and  European 
larch  in  the  whole  study  area,  and  of  height  variation  and 
mortality  causes  in  a  subsample.  Survival  of  each  tree 
was  recorded  annually  as  from  1976  (number  of  square 
units  with  cembran  pines  =  1,351,  with  mountain  pines  = 
1,350,  and  with  larches  =  1,351).  Mean  tree  height,  gen- 
eral constitution,  type  and  degree  of  abiotic  and  biotic 
damage,  and  causes  of  mortality  were  determined  annu- 
ally in  228  square  units  containing  cembran  pines,  in  226 
square  imits  containing  mountain  pines,  and  in  226 
square  units  containing  larches.  Twenty-eight  qualitative 
"types"  of  damage  and  potential  causes  of  death  could  be 
attributed  in  the  field. 

Correlation  analysis  was  used  to  test  for  interactions 
between  various  site  parameters  and  tree  performance. 
The  relative  importance  of  site  parameters  was  evaluated 
by  stepwise  regression  analyses  (SAS  1985)  for  survival 
and  tree  height.  The  survival  rates  were  arcsine  trans- 
formed before  performing  stepwise  regression  analysis. 

RESULTS 
Survival 

During  the  first  3  years  after  planting,  survival  was 
highest  in  cembran  pine  compared  to  mountain  pine  and 
larch,  but  in  1991  only  25.3  percent  of  the  planted  cembran 
pines  survived,  compared  to  37.8  percent  in  mountain 
pine  and  73.1  percent  in  larch  (fig.  2). 

Stepwise  regression  of  survival  rate  in  a  square  unit 
(number  of  trees  alive  in  1991/number  of  trees  planted  in 
1975)  of  cembran  pine  revealed  that  the  date  of  disappear- 
ance of  snow  cover  in  spring  and  the  altitude  above  sea 
level  were  the  most  important  factors,  explaining  some 
46  percent  and  8  percent  of  total  variation,  respectively. 
In  our  study  area  the  survival  rate  was  high  on  the  early 
snow-free  and  lower  sites.  Survival  decreased  with  in- 
creasing altitude,  and  almost  no  trees  survived  on  sites 
where  snow  lasted  on  10-year  average  longer  than  June  10. 


Avalanche  frequency,  slope  inclination,  global  radiation, 
and  the  number  of  snow-free  days  in  winter  also  had  sig- 
nificant effects  on  survival,  but  each  of  these  factors  con- 
tributed less  than  1  percent  to  the  total  variation. 

Mortality  Causes 

Of  the  28  recorded  causes  for  mortality,  Ascocalyx 
able-  tina  (Lagerb.)  (=  Gremmeniella  abietina  [Lagerb.] 
Morelet),  a  parasitic  fungus,  was  the  major  factor,  killing 
some  52.3  percent  of  the  planted  cembran  pines.  A  second 
fungus,  Phacidium  infestans  Karst.,  the  snow  blight, 
killed  another  11.8  percent  of  the  trees.  Some  4.8  percent 
of  the  trees  died  or  disappeared  without  any  detectable 
reason.  Undetermined  fungi  killed  1.3  percent  and  com- 
petition with  herbaceous  vegetation  some  0.6  percent  of 
the  planted  trees.  Herbivores  caused  no  losses. 

We  found  a  significantly  positive  correlation  between  the 
proportion  of  cembran  pines  killed  by  Ascocalyx  abietina 
and  the  date  of  disappearance  of  snow  cover  in  spring 
(fig.  3).  Square  units  that  were  snow  free  before  May  12 
suffered  only  slightly  (13.9  percent  of  the  trees  killed), 
whereas  AscocaZyx  killed  90.3  percent  of  the  trees  in  units 
that  were  covered  by  snow  later  than  June  10.  Occur- 
rence oi Ascocalyx  was  also  positively  correlated  with 
altitude  {r  =  0.55,  P  <  0.0001,  n  =  228),  the  fungus  kilHng 
an  increasing  proportion  of  yoimg  trees  with  increasing 
altitude. 

Between  the  proportion  of  cembran  pines  killed  by  Pha- 
cidium infestans  and  the  date  of  disappearance  of  snow 
cover  in  spring,  however,  we  found  a  negative  correlation 
(r  =  -0.42,  P  <  0.001,  n  =  228).  This  means  that  the  high- 
est proportion  of  trees  was  killed  in  squares  that  were 


75  76  77  78  79  80  81  82  83  84  85  86  87  88  89  90  91 


Year 

Figure  2— Survivorship  curves  of  the  three 
tree  species,  cembran  pine  (Pinus  cembra), 
mountain  pine  (P.  uncinata),  and  European 
larch  {Larix  decidua),  planted  in  1975  in  the 
Stillberg  experimental  area. 


106 


1.0 


p 

E  o 

O  0) 

a 
o 


0.8  • 


0.6 


0.4 


0.2 


•  .  .  •  •  • 


•  •  •       I  • 


•i 

•t  . 


•     •  • 

t'..'  «. 


120  130  140  150  160  170 

Date  of  Disappearance  of  Snow  Cover  (day) 

Figure  3 — Relationship  between  the  propor- 
tion of  cembran  pines  killed  in  a  square  unit  by 
Ascocalyx  abietina  (=  number  of  trees  killed  by 
Ascocalyx  since  1 976/number  of  trees  planted 
in  1975)  and  the  date  of  disappearance  of  snow 
cover  in  spring.  Day  1 30  =  May  1 0;  day  1 60  = 
June  9.  Regression  line:  y  =  -3.0637  +  0.0251  x, 
r=0.8^,P«  0.0001,  n  =  228. 


snow  free  early,  whereas  the  importance  of  Phacidium 
as  a  mortaHty  cause  decreased  with  increasing  duration 
of  snow  cover.  Similarly  the  ratio  of  trees  killed  by  Pha- 
cidium decreased  with  increasing  altitude  {r  =  —0.26, 
P<0.0001,  n  =  228). 

The  occurrences  of  the  two  parasitic  fungi  were  signifi- 
cantly negatively  correlated  within  square  units  (r  =  -0.59, 
P  <  0.0001,  n  =  228).  If  one  of  the  two  fungus  species  was 
common  in  one  square,  the  other  was  rare. 


DISCUSSION 
Survival 

Duration  of  snow  cover  in  spring  was  found  to  be  the 
main  limiting  factor  for  survival  of  young  cembran  pines 
in  the  upper  subalpine  forest  belt.  On  sites  where  snow 
cover  lasted  until  after  Jime  10  cembran  pines  were  al- 
most completely  excluded.  The  second  evergreen  Pinus 
species,  the  mountain  pine,  was  similarly  vulnerable  to 
long-lasting  snow  cover.  The  European  larch,  a  deciduous 
conifer,  was  much  more  tolerant.  In  this  species  30  per- 
cent of  the  trees  siu^ved  on  sites  snow  covered  until  after 
June  10.  Further,  survival  of  cembran  pine  was  signifi- 
cantly affected  by  altitude,  although  this  factor  was  clearly 
less  important  than  duration  of  snow  cover  in  spring.  In- 
terestingly, the  impact  of  site  factors  such  as  global  radia- 
tion, number  of  snow-free  days,  slope  inclination,  and  ava- 
lanche frequency  on  tree  survival  was  negligible  compared 
to  the  first  two  factors.  Avalanche  frequency,  however, 
will  certainly  become  more  important  in  the  near  future. 
Young  trees  with  flexible  stems  generally  survive  ava- 
lanche impacts  without  significant  damage.  But  with  in- 
creasing height  and  stem  diameter  the  trees  are  becoming 
increasingly  vulnerable  to  stem  breakage  caused  by  snow 
movements,  a  tendency  that  has  become  evident  only  dur- 
ing the  last  years. 

Mortality  Causes 

In  our  study  area  the  two  parasitic  fungi  Ascocalyx 
abietina  and  Phacidium  infestans  were  the  major  mortal- 
ity factors  for  cembran  pines,  and  these  fungi  still  kill  sig- 
nificant numbers  of  young  trees  every  year. 


Tree  Height 

Mean  tree  height  in  1991  was  lowest  in  cembran  pine 
(58.8  cm)  compared  to  moxmtain  pine  (72.7  cm)  and  larch 
(59.8  cm),  although  initial  mean  height  was  largest  in 
cembran  pine  (fig.  4). 

Variation  in  height  was  considerable,  depending  on 
local  site  conditions.  On  sunny  southeast  exposed  sites, 
average  height  of  cembran  pines  was  above  80  cm,  whereas 
on  shady  sites  mean  height  was  around  45  cm.  Stepwise 
regression  revealed  that  tree  height  was  mainly  affected 
by  altitude,  global  radiation,  and  wind  velocity,  these  fac- 
tors contributing  21  percent,  15  percent,  and  10  percent 
to  the  total  variation.  Duration  of  snow  cover  in  spring 
explained  3  percent  of  the  variation  in  height.  In  our 
study  area  the  cembran  pines  were  high  on  sites  at  low 
altitudes,  receiving  high  amounts  of  radiation  and  being 
protected  from  strong  winds.  Slope  inclination,  avalanche 
frequency,  and  number  of  snow-free  days  in  winter  had  no 
significant  impact  on  tree  height. 


E 

o 


c 

(0 
0) 


80 


70 


60 


SO 


X 


30 


20 


10 


1 

-♦■  - 

-  C€ 

-  M< 

-  La 

>mb 
}unt 
rch 

'an  pine 
ain  pine 

r' 

r' 

— 

-«-•• 

r' 

,  

✓ 

r' 

r'' 

b-' 



✓ 

...I 

p: — 

r' 

— yt 
(' 

1  1 



— 

r-- 

r-- 

r 

..-1 

r' 

r'' 

r 



1  .- 

■ 

75  76  77  78  79  80  81  82  83  84  85  86  87  88  89  90  91 

Year 

Figure  4 — Annual  mean  height  of  the  three  tree 
species,  cembran  pine  {Pinus  cembra),  mountain 
pine  (P.  uncinata),  and  European  larch  (Larix  de- 
cidua),  planted  in  1975  in  the  Stillberg  experimen- 
tal area. 


107 


Ascocalyx  infestations  first  become  evident  in  spring 
in  the  dying  buds.  Later  in  the  season  the  shoots  die  back 
along  the  branches.  The  fungus  may  eventually  kill  a 
young  cembran  pine  after  2  or  3  years.  The  close  relation- 
ship between  the  date  of  disappearance  of  snow  cover  in 
spring  and  the  occurrence    Ascocalyx  (fig.  3)  indicates 
that  the  fungus  preferentially  attacks  weakened  plants. 
On  the  sites  with  long-lasting  snow  cover,  the  vegetation 
season  is  relatively  short,  and  thick  layers  of  raw  himius 
have  accumulated  because  of  the  low  temperatures. 
Therefore,  they  are  poor  sites  for  tree  growth.  Similarly, 
Kurkela  (1984)  found  that  the  trees  most  heavily  affected 
hy  Ascocalyx  had  the  lowest  growth  rate  before  the  epi- 
demic. He  explained  the  differences  in  tree  growth  by  the 
different  site  fertility.  In  our  study  the  hypothesis  that 
Ascocalyx  preferentially  attacked  weakened  trees  was  fur- 
ther supported  by  the  finding  that  the  proportion  of  trees 
killed  by  the  fungus  significantly  increased  with  increas- 
ing altitude  (toward  the  altitudinal  tree  line).  In  our  study 
area  Ascocalyx  attacked  mountain  pines,  too,  where  it  also 
caused  considerable  losses. 

In  contrast  to  the  situation  in  Ascocalyx,  the  proportion 
of  cembran  pines  killed  by  Phacidium  was  negatively  cor- 
related with  the  date  of  disappearance  of  snow  cover  and 
with  the  altitude.  According  to  this  correlation  the  trees 
growing  at  lower  altitudes  and  on  sites  that  were  snow 
free  early  were  killed  more  frequently  by  Phacidium  than 
the  trees  growing  at  higher  altitudes  and  on  sites  with 
long-lasting  snow  cover.  Young  trees  that  looked  healthy 
in  one  year  were  often  found  dead  in  the  following  spring. 
Interestingly,  Roll-Hansen  and  others  (1992)  reported 
from  Scandinavia  that  the  most  vigorously  growing  plants 
were  most  susceptible  to  Phacidium  infestations,  whereas 
poorly  developing  plants  were  most  resistant. 

The  fact  that  Phacidium  killed  the  highest  proportion 
of  trees  at  early  snow-free  sites  seems  contradictory  to 
the  biology  of  the  fungus,  that  exclusively  attacks  parts 
of  trees  that  are  covered  by  snow  (Roll-Hansen  1989). 
Branches  destroyed  by  this  fungus  are  clearly  visible  after 
disappearance  of  snow  cover  and  mark  the  prevalent  snow 
depth.  Our  study  area  contains  some  older  cembran  pines 
growing  on  ridges  (at  the  sites  that  are  snow  free  early). 
These  trees  originating  from  caches  of  the  nutcracker 
{Nucifraga  caryocatactes)  host  Phacidium,  which  is  non- 
lethal  for  the  large  trees  that  extend  considerably  beyond 
the  snow  cover.  These  trees  may  act  as  sources  for  local 
Phacidium  outbreaks  that  eventually  kill  the  young  trees 
completely  covered  by  snow.  Since  this  fungus  specifically 
attacks  cembran  pine  and  does  not  occur  on  mountain 
pine  (but  see  Roll-Hansen  1989),  it  may  not  spread  as  eas- 
ily over  larger  areas  as  does  Ascocalyx,  which  uses  moun- 
tain pine  as  an  alternative  host.  Therefore,  even  heavier 
outbreaks  of  Phacidium  remain  locally  restricted  around 
mature  cembran  pines.  Together  with  the  fact  that  Pha- 
cidium preferentially  attacks  the  most  vigorous  trees,  this 
may  explain  the  seemingly  paradoxical  negative  correla- 
tion between  duration  of  snow  cover  in  spring  and  occur- 
rence o{  Phacidium. 

On  sites  with  high  avalanche  frequencies  the  young 
trees  were  partially  excluded,  not  by  stem  breakage  or 
uprooting,  but  by  competitive  interactions  with  herba- 
ceous vegetation.  The  moving  snow  carried  rich  mineral 


soil  leading  to  a  lush  herbaceous  vegetation  that  out- 
competed  the  slowly  growing  young  cembran  pines.  Ava- 
lanches, however,  become  a  more  serious  problem  for  the 
surviving  trees  as  they  increase  in  height  and  stem  diam- 
eter. Broken  stems  were  found  more  frequently  in  recent 
years. 

Herbivore  impact  on  the  survival  of  young  cembran  pines 
was  relatively  unimportant.  Herbivores  killed  no  trees,  al- 
though they  may  reduce  tree  growth.  Browsing  vertebrates 
like  black  grouse  (Tetrao  tetrix)  and  chamois  (Rupicapra 
rupicapra),  however,  clearly  preferred  mountain  pine  and 
larch  to  cembran  pine.  Black  grouse  caused  some  locally 
restricted  damage  by  browsing  buds  and  first-year  needles. 
They  obviously  preferred  to  browse  on  easily  accessible 
trees  growing  on  ridges  with  little  snow  cover  (Streule 
1973).  Populations  of  herbivorous  insects  like  phloem- 
sucking  aphids  (Cinara  cembrae  Seitner,  see  Grbic,  these 
proceedings,  and  Pineus  cembrae  [Choi.]  Amand.)  fluctu- 
ated in  size  but  were  low  in  most  of  the  years  when  com- 
pared to  the  other  two  tree  species.  Although  visible  dam- 
age by  herbivores  was  relatively  rare,  much  rarer  than 
fungal  infections,  the  herbivores  may  have  indirectly  af- 
fected the  survival  of  the  trees  by  injuring  plant  tissue, 
making  the  trees  more  susceptible  to  infestations  by  fungi. 

Tree  Height 

Tree  height  in  cembran  pine  was  clearly  dependent  on 
the  available  amount  of  energy  during  the  vegetation  sea- 
son. Altitude  above  sea  level,  global  radiation,  and  wind 
velocity  explained  the  largest  amount  of  variation  in 
height.  We  are  aware  that  tree  height  does  not  directly 
represent  tree  growth,  since  we  did  not  measure  annual 
growth.  We  only  recorded  tree  height  in  summer,  exclud- 
ing the  growing  shoot  of  the  current  year.  Tree  height, 
which  is  the  result  of  annual  growth  minus  the  losses 
through  damage  in  the  apical  region,  may  increase  or 
decrease  from  one  year  to  another.  Sixteen  years  after 
plantation,  differences  in  average  tree  height  among  sites, 
however,  are  primarily  the  result  of  variation  in  growth 
caused  by  variation  in  local  site  conditions. 

The  date  of  disappearance  of  snow  cover  in  spring,  al- 
though significantly  related  to  tree  height,  explained  only 
3  percent  of  the  variation  in  height  in  contrast  to  survival 
where  the  snow  conditions  in  spring  mainly  explained  the 
observed  pattern. 

Slope  inclination  and  avalanche  frequency,  which  may 
be  related  to  stem  breakage,  had  no  detectable  impact  on 
tree  height.  Further,  we  found  no  relationship  between 
the  number  of  snow-free  days  in  winter  and  tree  height, 
although  trees  that  were  snow  free  during  extended  peri- 
ods in  winter  should  have  been  more  vulnerable  to  herbi- 
vore browsing  than  trees  that  were  covered  by  deep  snow. 

CONCLUSIONS 

Foresters  should  carefully  select  favorable  microsites 
when  planting  cembran  pines  in  high-altitude  afforestations. 
Particular  attention  should  be  paid  to  the  spatial  pattern 
of  snowmelt  in  spring,  and  no  cembran  pines  should  be 
planted  at  sites  with  prolonged  duration  of  snow  cover. 


108 


REFERENCES 

Blaser,  P.  1980.  Der  Boden  als  Standortsfaktor  bei 
Aufforstiingen  in  der  subalpinen  Stufe  (Stillberg, 
Davos).  Eidgenossische  Anstalt  fur  das  forstliche 
Versuchswesen,  Mitteilimgen.  56(3):  527-611. 

Kuoch,  R.  1970.  Die  Vegetation  auf  Stillberg  (Dischmatal, 
Kt.  Graubiinden).  Eidgenossische  Anstalt  fiir  das 
forstliche  Versuchswesen,  Mitteilungen.  46(4):  329-342. 

Kurkela,  T.  1984.  The  growth  of  trees  affected  by  Grem- 
meniella  abietina.  In:  Manion,  P.  D.,  ed.  Scleroderris 
canker  of  conifers.  The  Hague:  Martinus  NijhofiEDr  W. 
Junk  Pubhshers:  177-180. 

Nageh,  W.  1971.  Der  Wind  als  Standortsfaktor  bei 
Aufforstungen  in  der  subalpinen  Stufe  (Stillbergalp 
im  Dischmatal,  Kanton  Graubiinden).  Eidgenossische 
Anstalt  fiir  das  forstliche  Versuchswesen,  Mitteilungen. 
47(2):  33-147. 

Roll-Hansen,  F.  1989.  Phacidium  infestans:  a  literature 
review.  European  Journal  of  Forest  Pathology.  19: 
237-250. 

Roll-Hansen,  F.;  Roll-Hansen,  H.;  Skroppa,  T.  1992. 
Gremmeniella  abietina,  Phacidium  infestans,  and  other 
causes  of  damage  in  alpine,  young  pine  plantations  in 
Norway.  European  Journal  of  Forest  Pathology.  22: 
77-94. 

Rychetnik,  J.  1987.  Snow  cover  disappearance  as  influ- 
enced by  site  conditions,  snow  distribution,  and  ava- 
lanche activity.  In:  Proceedings  of  the  international 
symposiima  on  avalanche  formation,  movement  and 
effects;  Davos  1986.  lAHS-Pubhcation  Nr.  162. 

SAS.  1985.  SAS  user's  guide:  statistics.  Gary,  NC:  SAS 
Institute. 


Schonenberger,  W.  1975.  Standorteinfliisse  auf 
Versuchsaufforstimgen  an  der  alpinen  Waldgrenze 
(Stillberg,  Davos).  Eidgenossische  Anstalt  fiir  das 
forstliche  Versuchswesen,  Mitteilungen.  51(4):  357-428. 

Schonenberger,  W.  1985.  Performance  of  a  high  altitude 
afforestation  under  various  site  conditions.  In:  T\irner, 
H.;  Tranquillini,  W.,  eds.  Establishment  and  tending  of 
subalpine  forest:  research  and  management:  Proceed- 
ings of  the  3rd  lUFRO  workshop  P  1.07-00,  1984. 
Eidgenossische  Anstalt  fiir  das  forstUche  Versuchswesen, 
Berichte.  270:  233-240. 

Schonenberger,  W.;  Frey,  W.,  eds.  1988.  Untersuchimgen 
zur  Okologie  und  Technik  der  Hochlagenaufforstung; 
Forschungsergebnisse  aus  dem  Lawinenanrissgebiet 
Stillberg.  Schweizerische  Zeitschrift  fiir  Forstwesen. 
139:  735-820. 

Schonenberger,  W.;  Frey,  W.;  Leuenberger,  F.  1990. 
Okologie  und  Technik  der  Aufforstung  im  Gebirge  - 
Anregimgen  fiir  die  Praxis.  Eidgenossische  Anstalt  fiir 
das  forstliche  Versuchswesen,  Berichte  Nr.  325.  58  p. 
[Version  franfaise:  Ecologie  et  technique  des  afforesta- 
tions en  montagne  -  Suggestions  a  I'usage  des  praticiens. 
Traduction  du  rapport  No.  325.]  [Versione  italiana: 
Ecologia  del  rimboschimento  di  montagna  e  pratica 
della  tecnica  di  piantagione.  Traduzione  del  rapporto 
No.  325.] 

Streule,  A.  1973.  Schaden  in  Gebirgsaufforstungen  durch 
das  Birkhuhn  (Lyrurus  tetrix).  Biindnerwald.  28(8): 
249-254. 

Tiimer,  H.  1966.  Die  globale  Hangbestrahlung  als 
Standortsfaktor  in  der  subalpine  St\ife.  Eidgenossische 
Anstalt  fiir  das  forstliche  Versuchswesen,  Mitteilungen. 
42(3):  111-168. 


109 


Influence  of 
Environmental  Factors 


International  Workshop 
St.  Moritz  1 992 

s   ^' 


110 


CHANGES  OF  SWISS  STONE  PINE 
APHID  LIFE  CYCLE,  DENSITY,  AND 
POPULATION  STRUCTURE  IN  HIGH- 
ALTITUDE  SWISS  STONE  PINE 
AFFORESTATION 

Mihailo  Grbic 


Abstract — The  order  of  form  appearance,  structure,  and  number 
of  individuals  in  colonies  of  Cinara  cembrae  on  Swiss  stone  pine 
{Pinus  cembra)  used  for  afforestation  in  the  Dischma  Valley  were 
strongly  related  to  the  slope  aspect.  The  abnormality  was  par- 
ticularly apparent  on  north  slopes.   North  slope  colonies 
showed  incomplete  life  cycles.  All  insects  seen  there  were 
virginoparae  and  their  larvae.  Colonies  were  small  in  number 
and  were  developed  from  winged  virginoparae  from  other  areas. 


The  Alps  are  of  great  importance  to  Switzerland  as  an 
area  for  hviman  habitation  and  as  a  health  and  holiday  re- 
sort. Afforestation  in  high  altitudes  is  costly  and  involves 
high  risk,  as  many  failures  occur.  Because  of  the  slow  de- 
velopment of  such  afforestation  it  is  difficult  to  under- 
stand and  to  determine  the  causes  and  processes  that  lead 
to  a  failure  and  then  to  suggest  measures  to  improve 
success. 

Environmental  factors  of  the  upper  timberline  area  are 
totally  different  from  those  of  the  forest  zones.  These 
harsh  conditions  affect  the  specific  growth  and  develop- 
ment of  planted  species;  for  example,  low  survival  rate 
and  low  annual  increment.  The  surviving  plants,  depend- 
ing on  the  biological  characteristics  of  the  species,  are 
more  or  less  open  to  insect  and  fungal  attack.  These  same 
high-altitude  factors  (for  example,  low  temperature,  short 
growing  period,  and  short  day  photoperiod)  tend  to  change 
the  life  cycle,  density,  and  population  structure  of  insects. 
This  is  particularly  true  for  aphids  due  to  their  polymor- 
phic nature,  and  their  ability  to  alter  the  nimaber  of  gen- 
erations per  season. 

STUDY  AREAS  AND  METHODS 

These  investigations  were  carried  out  in  field  plots  of 
the  Stillberg  research  area  and  the  Lucksalp  comparative 
afforestation  area,  in  the  timberline  area  of  the  Dischma 
Valley  (Canton  Graubunden).  The  period  of  the  study 
was  between  late  May  and  the  middle  of  September, 
which  is  the  time  of  greatest  insect  activity. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Mihailo  Grbic  is  Teacher  at  the  University  of  Belgrade,  Faculty  of  For- 
estry, Kneza  Viseslava  1,  YU-11030  Belgrade,  Yugoslavia. 


Stillberg  Hes  at  2,080  to  2,230  m  above  sea  level  (a.s.l.), 
with  a  northeast  aspect,  and  30°  to  45°  slope.  The  slope  is 
divided  by  spurs,  and  as  a  result  there  are  three  different 
aspects  (north-,  northeast-,  and  east-facing  slopes).  On 
the  opposite  side  of  Dischma  Valley  lies  Lucksalp  at 
2,200  m  a.s.l.,  with  a  southwest  aspect. 

The  subjects  of  investigation  were  19-year-old  Swiss 
stone  pine  (Pinus  cembra  L.)  and  the  Swiss  stone  pine 
aphid  (Cinara  cembrae  Seithner),  the  most  frequently  ob- 
served insect  pest  on  that  afforestation  species.  The  trees 
were  planted  70  cm  apart  in  a  grid  pattern  in  approxi- 
mately 4,000  square  plots  (3.5  by  3.5  m).  The  Swiss  stone 
pine  stocks  were  alternated  from  plot  to  plot  with  two 
other  afforestation  species,  17-year-old  Swiss  mountain 
pine  (Pinus  mugo  ssp.  uncinata  [Mill.]  Domin.),  and 
15-year-old  European  larch  (Larix  decidua  Mill.),  each 
plot  containing  25  trees  (Shonenberger  1985). 

The  samples  of  Swiss  stone  pine  aphid  were  identified 
by  comparing  them  to  recent  accounts  from  various  parts 
of  Europe  (Carter  and  Maslen  1982;  Eastop  1972;  Pintera 
1966;  Stroyan  1955,  1960).  The  structure  of  antennae, 
rostrum,  abdomen,  and  hind  legs  of  collected  material 
were  compared  with  these  references. 

The  life  cycle  was  observed  in  colonies  feeding  on  par- 
ticular sample  trees.  On  Lucksalp,  30  trees  (5  trees  x 
6  plots)  were  regtdarly  examined,  and  50  (5  trees  x 
10  plots)  were  examined  on  Stillberg.  Observations  were 
made  to  establish  the  relationship  between  the  intensity 
of  infestation  and  site  aspects. 

The  trees  were  examined  every  10  days  to  determine 
density  of  pest  population  and  structvire  of  colonies. 

MORPHOLOGY 

The  morphological  characteristics  of  specimens  found 
on  Stillberg  and  Lucksalp  (tables  1  and  2)  are  similar  to 
the  central  European  description  (Pintera  1966).  The 
chronological  series  of  morphs  is:  fundatrix,  aptera 
virginopara,  alata  virginopara,  ovipara,  and  aptera  and 
alata  males. 

Compared  with  the  apterae  of  the  succeeding  genera- 
tions, fundatrix  shows  "fundatrix  facies"  characteristics 
(Lees  1961).  With  the  Swiss  stone  pine  aphid,  the  anten- 
nae and  legs  are  relatively  shorter  in  comparison  to  the 
body.  The  processus  terminalis  is  shorter  with  fewer 
rhinaria.  Features  similar  to  the  fundatrix  facies  that 
were  observed  by  Hille  Ris  Lambers  (1955),  Lees  (1961), 
and  Stroyan  (1960)  on  nonfimdatrix  morphs  of  some 


111 


Table  1 — Biometric  data  for  different  morphs  of  Cinara  cembrae  from  Lucksalp 


Biometric  data 


Aptarae 
virginoparae 


Alatea 
virginoparae 


Morph 
Apterae 
male 


Alatae 
male 


Sexual 
female 


Body  length  (mm) 

Length  of  antennal  segments 
II 
IV 
V 

VI  (base  +  proces.  termlnalls) 

Length  of  rostal  segments  (^) 
IV 
V 

Hind  tarsus  segments  (^i) 
I 

basal  diameter 
dorsal  length 
ventral  length 

II 

length 

Hind  tibia  length  (mm) 

Length  of  longest  hair  on 
III  ant.  segment 
hind  tibia 

VIII  abdom.  tergite 

Number  of  hairs  on 
II  ant.  segment 
VI  ant.  segment 
base 

proces.  term, 
ultimate  rostral  segment 

Number  of  hairs  on 
subgenital  plate 
VIII  abdom.  tergite 

Number  of  secondary  rhinaria  on 
ant.  segments 

III 

IV 

VI 


3.05-3.52 


680-720 
300-340 
400-410 
167-180  + 
75-83 

218-238 
100-112 


74-  78 
95-105 

155-182 

385-420 
3.20-3.30 

75-  97 
75-120 

110-165 

9-15 

9-  13 
4-5 
6-8 

10-  12 
4-6 


0-2 
0-2 
2-3 


3.00-3.10 


635-650 
275-290 
320-345 
150-160  + 
59-66 

170-182 
85-88 


57-62 
62-66 
122-130 

330-350 
2.90-3.05 

88-92 
70-105 
185-200 

6-7 

10-12 
3-4 
8 

10 


9-11 
2-3 
5 


2.50-3.30 


830-880 
335-400 
430-500 
190-198  + 
75-90 

230-250 
90-110 


60-70 
98-110 
150-180 

340-390 
1.85-3.00 

75-120 
75-110 
100-112 

10-15 

10-12 
5 

10-12 

10-12 
7-11 


10-13 
7-10 
4-5 


2.40-3.00 


845-910 
360-430 
440-500 
109-198  + 
85-95 

220-230 
105-110 


60-65 
85-105 
150-180 

385-400 
2.90-3.40 

115-140 
140-180 
90-120 

7-  14 

8-  10 

5-  6 
8-10 

6-  9 
6-9 


55-83 
10-13 
5 


3.80-4.70 


830-890 
360-385 
480-500 
192-212  + 
75-80 

245-250 
106-118 


68-88 
105-120 
187-200 

370-420 
3.70-3.78 

95-123 
112-145 
120-138 

10-16 

8-  13 
5-6 

9-  11 

9-13 
7-8 


2-3 
1-2 
4-7 


aphids  that  live  at  low  temperatures  were  not  recorded  in 
these  study  areas. 

Morphological  differences  between  apterous  and  alatae 
virginoparae  are  found  not  only  in  the  presence  or  ab- 
sence of  the  wings,  but  also  in  the  following  characteris- 
tics. In  alatae: 

•  Segmentation  between  head  and  prothorEix  is  more 
conspicuous. 

•  Compound  eyes  are  bigger,  v(ath  many  facets. 

•  Antennae  bear  five  times  as  many  rhinaria. 

•  The  body  is  shorter,  and  the  mesothoracic  terga  con- 
sists of  a  compound  notum  and  postnotum. 

•  The  scutum  occupies  the  central  major  part  of  the 
nottun  and  is  divided  in  two  mesothoracic  lobes,  which  are 
hardly  developed. 


The  virginoparae  of  every  generation  are  morphologi- 
cally uniform,  and  under  field  conditions  it  is  impossible 
to  find  out  how  many  generations  have  developed.  Also, 
the  length  of  time  required  for  passing  through  four  in- 
stars,  from  birth  to  adult,  is  variable  and  dependent  on 
two  external  factors  (food  quality  and  temperature)  and 
two  internal  factors  (birth  weight  and  whether  the  morph 
is  winged  or  unwdnged)  (Dixon  1987).  This  is  the  reason 
for  the  presence  of  different  generations  of  the  same  form 
occurring  simultaneously  in  one  colony. 

As  is  known,  Swiss  stone  pine  aphids  produce  both 
alatae  and  apterous  males.  Both  forms  were  found  in  the 
research  areas,  but  with  a  preponderance  of  the  wingless 
form.  Both  forms  of  males  are  readily  recognized  by  their 
sclerotized  genital  structure.  The  body,  especially  abdo- 
men, is  smaller  and  more  slender,  and  the  antennae  bear 
more  rhinaria. 


112 


Table  2 — Biometric  data  for  different  morphs  of  Cinara  cembrae  from  Stillberg 


Morph 


Aptarae 

Alatea 

Apterae 

Alatae 

Sexual 

Biometric  data 

virginoparae 

virginoparae 

male 

male 

female 

Body  length  (mm) 

2.90-3.70 

2.90-3.90 

2.50-3.10 

2.40-3.00 

3.90-4.70 

Length  of  antennal  segments 

II 
II 

540-800 

^  Art    rt  C  rt 

630-650 

820-800 

840-930 

800-860 

IV 

260-355 

250-280 

320-400 

360-440 

350-400 

V 

340-450 

www  OOU 

tOU-wuU 

ARn-t^  1  n 

HOU  0  1  u 

u-ouu 

VI  (base  +  proces.  termlnalis) 

170-188-1- 

153-167  + 

183-210  + 

192-200  + 

175-222  + 

80-85 

58-68 

75-95 

88-95 

75-85 

Length  of  rostal  segments  (n) 

IV 

rtOQ  OCA 

238-250 

1 62-1 72 

235-250 

220-225 

235-250 

V 

100-108 

83-87 

95-102 

100-105 

105-118 

Hind  tarsus  segments  (n) 
1 

h^^^sl  dif^mptAr 

75-78 

WW  w9 

DU-DO 

DO- 1 UO 

dorsal  length 

88-108 

60-62 

95-1 02 

88-1 07 

105-125 

ventral  length 

II 

158-175 

122-127 

150-178 

150-170 

185-230 

length 

365-385 

300-340 

335-395 

380-400 

420-450 

Hind  tibia  length  (mm) 

2.40-3.39 

2.82-2.98 

1 .78-2.98 

2.95-3.37 

3.70-3.88 

Length  of  longest  hair  on 

III  ant.  segment 

60-00 

85-88 

78-125 

1 1 2-1 42 

95-130 

hind  tibia 

60-105 

75-1 10 

78-1 12 

140-175 

1 12-150 

VIII  aooom.  tergite 

168-220 

195-210 

100-115 

88-130 

112-138 

Number  of  hairs  on 

II  ant.  segment 

11-15 

6-7 

10-16 

7-15 

10-18 

VI  ant.  segment 

base 

8-12 

9-13 

10-13 

8-1 1 

8-16 

proces.  term. 

4-5 

3-4 

5-6 

5 

4-6 

ultimate  rostral  segment 

8-9 

7-9 

9-12 

8-11 

8-11 

Number  of  hairs  on 

subgenital  plate 

9-11 

6 

9-12 

6-9 

9-14 

VIII  abdom.  tergite 

4-6 

7-12 

6-9 

6-9 

Number  of  secondary  rhinaria  on 

ant.  segments 

III 

0-4 

8-10 

10-15 

50-89 

2-4 

IV 

2-3 

1-3 

7-11 

11-15 

0-2 

VI 

3-4 

6 

4-6 

5-6 

2-7 

Alatae  and  apterous  males  and  alatae  virginoparae  pos- 
sess a  far  larger  number  of  secondary  rhinaria  than  apter- 
ous virginoparae  and  sexuales  females.  The  following 
numbers  of  secondary  rhinaria  on  the  third  antennal  seg- 
ment were  found:  apterous  virginoparae  0-4,  alatae 
virginoparae  8-11,  apterous  males  10-15,  and  alatae 
males  50-89.  The  number  of  secondary  rhinaria  on 
sexuales  females  was  similar  to  that  found  in  apterous 
virginoparae. 

The  function  of  secondary  rhinaria  is  probably  different 
in  male  and  female  individuals.  The  secondary  rhinaria 
in  males  appear  to  be  the  main  receptors  of  sex  phero- 
mones,  while  secondary  rhinaria  of  alatae  virginoparae 
must  have  a  function  other  than  pheromone  detection. 
This  function  is  possibly  host  selection,  as  has  been  shown 
by  many  authors  with  other  aphid  species  (Alikhan  1960; 
Chapman  and  others  1981;  and  Pospisil  1976). 


Sexuales  females  resemble  the  apterous  virginoparae, 
but  may  be  somewhat  bigger.  The  tibiae  of  the  hind  legs 
are  often  longer  with  longer  setae  also,  and  numerovis 
pseudosensoria  on  them. 

LIFE  CYCLE 

General — The  life  cycle  of  the  Swiss  stone  pine  aphid  is 
strongly  related  to  site  aspects  and  environmental  factors. 
Generally,  the  Swiss  stone  pine  aphid  has  a  monoecious 
holocyclic  type  of  life  cycle.  The  life  cycle  starts  in  May 
v^dth  the  hatching  of  fundatrices  larvae  from  eggs  laid  in 
rows  on  the  needles.  In  the  covirse  of  svmimer  a  small 
niunber  of  generations  of  parthenogenetic  females 
develop. 

Generations  of  winged  parthenogenetic  viviparous  fe- 
males arise  from  colonies  of  apterae  later  in  the  summer. 


113 


Sexuales  forms  appear  in  the  second  half  of  August.  This 
early  development  of  sexuales  is  due  to  the  alpine  climate. 
In  the  low-lying  positions  where  Swiss  stone  pine  has 
been  artificially  introduced,  Swiss  stone  pine  aphid  has 
never  been  found. 

Observed — The  life  cycle  observations  in  the  Dischma 
Valley  show  the  following  characteristics:  colonies  from 
Lucksalp  and  all  slopes  of  Stillberg,  except  north-facing 
slopes,  show  complete  life  cycles  with  six  morphs.  Par- 
allel appearance  of  two  forms  of  males,  observed  by 
Seithner  (1936),  is  a  unique  case  in  the  family  Lachnidae. 
The  north-slope  colonies  showed  incomplete  life  cycles. 
All  insects  seen  there  were  virginoparae  apterae  or  alatae 
and  their  larvae.  Colonies  were  small  in  number  and 
were  developed  from  winged  virginoparae  from  other 
areas  that  colonized  plants  on  north-facing  slopes  from 
other  areas  (table  3). 

The  order  of  appearance  of  particular  forms  on 
Lucksalp  was  earlier  than  on  the  Stillberg  research  area. 
The  first  observation,  made  in  the  last  third  of  May, 
shows  68  percent  unhatched  eggs,  and  all  fundatrix  lar- 
vae were  part  of  the  first  two  larval  instars.  At  the  same 
time  on  Stillberg  87  to  90  percent  of  the  eggs  were 
unhatched. 

In  the  first  third  of  June  the  colony  structure  on 
Lucksalp  was:  7  percent  of  the  eggs  were  unhatched, 
young  fundatrices  in  all  four  larval  instars  were  found, 
but  no  adults.  A  similar,  but  slightly  different,  situation 
was  recorded  on  Stillberg.  There  were  25  to  38  percent 
unhatched  eggs,  and  the  colonies  contained  only  the  first 
three  larval  instars. 

During  the  observations  in  the  second  third  of  June  we 
found  adult  fundatrices  on  Lucksalp  and  on  the  east 
slopes  of  Stillberg.  With  the  appearance  of  fundatrices, 
the  density  of  populations  increased  gradually. 

On  the  other  Stillberg  slopes  (ENE  and  NE)  fundatrices 
were  found  10  days  later.  Simultaneously  we  recorded 
adults  of  apterous  virginoparae  on  Lucksalp.  This  is  a 
time  of  heavy  increases  in  population,  because  the  multi- 
plication rate  of  apterous  virginoparae  is  very  high,  and 
the  duration  of  larval  instars  very  short. 

In  an  experiment  that  was  conducted  under  conditions 
of  changeable  room  temperature,  one  apterous  virgino- 
parae was  isolated  on  a  twig  that  was  kept  moist.  Ten 
days  later,  we  recorded  six  apterous  virginoparae  of 
a  new  generation  (which  started  with  larviposition);  16 
in  the  fourth  larval  instar;  and  18  larvae  were  in  yoimger 
stadiums. 

It  was  presumed  that  the  effect  of  the  multiplication 
rate  is  not  the  same  in  nonisolated  field  conditions,  but 
on  some  trees  the  increments  of  population  for  10  days 
increased  by  a  factor  of  10  or  more.  On  the  other  hand, 
some  other  populations  were  substantially  reduced  by  the 
action  of  parasites  and  predators. 

In  the  first  third  of  July  at  Lucksalp  a  typical  popula- 
tion consisted  of  more  than  170  individuals;  almost  all 
belonged  to  apterous  virginoparae  larvae  (83  percent)  or 
adult  insects  ( 15  percent).  The  rest  (2  percent)  were 
alatoid  larvae  of  alatae  virginoparae.  The  fundatrix  gen- 
eration was  dead.  At  Stillberg,  populations  were  smaller 


with  most  of  the  population  made  up  of  larvae  and  adults 
of  apterous  virginoparae;  however,  fundatrices  were  also 
found  to  be  present.  The  fourth  larval  instar  of  alatae 
virginoparae  was  recorded  on  the  east  slopes  only.  The 
proportion  of  alatoid  larvae  was  10  percent  of  the  total 
number  of  the  fourth  larval  instar. 

The  first  Swiss  stone  pine  aphid  was  recorded  on  the 
north  slopes  of  Stillberg  in  the  middle  of  July.  Colonies 
were  small  and  consisted  of  only  a  few  adult  alatae 
virginoparae  and  young  larvae.  Few  of  the  colonies  were 
without  adults.  The  population  structures  of  the  other 
slopes  were  unchanged  in  content  until  the  last  third  of 
August.  Only  the  numbers  of  individuals  were  variable, 
depending  on  weather  conditions. 

The  rapid  decrease  in  number  of  individuals  at  the  be- 
ginning of  August  was  caused  by  low  temperatures  and 
snow.  The  greatest  changes  in  population  quantity  were 
recorded  on  north  slopes.  The  average  population  struc- 
ture in  the  course  of  July  and  August  was:  The  majority 
of  individuals  belonged  to  various  larval  instars,  and  the 
adult  part  of  the  population  consisted  of  apterae  and 
alatae  viviparous  parthenogenetic  females.  Average  ratio 
of  winged  and  wingless  forms  was  1:5.3  with  an  increas- 
ing tendency  to  winged  forms  in  colder  sites.  This  ten- 
dency is  due  to  the  strategy  of  the  species  to  avoid  expo- 
sure to  more  extreme  climatic  conditions  (by  colonizing 
another  host  plant  of  the  same  species,  but  in  better  envi- 
ronmental conditions).  Extremes  of  the  ratio  were  1:8  and 
1:2.  The  first  was  recorded  on  Lucksalp,  and  the  second 
on  Stillberg  north  slopes. 

The  highest  population  numbers  were  found  in  the  last 
third  of  July,  after  which  the  numbers  decreased  rapidly 
as  a  consequence  of  low  temperatures  at  the  beginning 
of  August.  Thereafter,  the  population  number  increased 
quickly  again  to  a  secondary  peak  at  the  end  of  August. 
Occurrence  of  the  last  generation  was  recorded  after  an- 
other cool  weather  period  at  the  end  of  August.  As  the 
overwintering  eggs  are  produced  sexually,  the  last  genera- 
tion of  the  year  must  contain  males  and  sexual  females, 
instead  of  the  parthenogenetic  forms  that  occur  during 
the  rest  of  the  year.  However,  sexuales  do  not  occur  un- 
der all  investigated  conditions.  On  the  north  slopes  of 
Stillberg,  for  example,  no  sexuales  forms  were  recorded. 
On  Lucksalp,  occurrence  of  sexuales  forms  was  about 
10  days  before  Stillberg. 

The  first  appearance  of  eggs  was  observed  at  the  begin- 
ning of  September.  Oviparous  females  move  away  from 
the  colony  to  the  top  part  of  current  shoots,  and  lay  over- 
wintering eggs  in  rows  on  needle  clusters.  Eggs  were 
brown  and  later  black  and  shiny  48  hours  after  being  laid. 
Average  number  of  eggs  in  the  row  along  one  needle  was 
8.2  and  the  extremes  were  16  and  1.  Maximum  egg  num- 
ber per  one  colony  was  316,  recorded  in  the  first  week  of 
September  at  Lucksalp.  The  same  colony  contained  eight 
sexuales  females.  Results  of  dissection  of  females  show 
12  ripe  eggs  in  the  ovariole  on  average.  This  was  smaller 
in  comparison  with  the  embryo  numbers  obtained  by  dis- 
section of  parthenogenetic  females  during  July,  which 
was  30. 


114 


Table  3 — Average  number  of  different  morphs  of  Cinara  cembrae  on  Swiss  stone  pine  sample  trees  from  last  third  of  May  to  second  third  of 
September 




Morph 

May 

June 

Julv 

August 

September 

3/3 

1/3 

2/3 

3/3 

1/3  2/3 

3/3 

1/3 

2/3 

3/3 

1/3 

2/3 

Lucksalp  (SW  slope) 

Eqqs 

5.6 

0.6 

60.0 

93.0 

Larvae 

2.6 

7.6 

35.0 

103.0 

148.0  178.0 

199.0 

68.0 

82.0 

124.0 

9.0 

11.0 

Fundatrices 

1.2 

1.5 

Apterae  virginoparae 

2.0 

27.0  12.0 

24.0 

8.3 

18.0 

24.0 

7.3 

6.7 

Aiatae  virginoparae 

1.5 

2.4 

1.8 

1.5 

2.0 

0.2 

Oviparous  females 

0.1 

4.9 

5.4 

Apterae  males 

0.1 

0.3 

1.4 

Aiatae  males 

0.1 

Stillberg  (E  slope) 

Eqqs 

7.2 

2.0 

0.2 

3.8 

57.0 

Larvae 

0.8 

4.4 

30.0 

41.0 

58.0  100.0 

110.0 

54.0 

80.0 

104.0 

5.9 

8.1 

Fundatrices 

0.6 

1.8 

1.0 

Apterae  virginoparae 

3.6  30.0 

31.0 

6.3 

11.0 

20.0 

2.9 

3.2 

Aiatae  virginoparae 

3.0 

5.0 

5.8 

0.7 

Oviparous  females 

0.2 

2.3 

Apterae  males 

0.1 

0.3 

Aiatae  males 

0.1 

0.2 

Stillberg  (ENE  slope) 

Eqqs 

7.0 

3.0 

1.1 

Larvae 

0.8 

4.8 

6.7 

34.0 

124.0  140.0 

145.0 

54.0 

69.0 

93.0 

4.6 

8.3 

Fundatrices 

2.4 

0.2 

Apterae  virginoparae 

5.0  12.0 

18.0 

3.2 

10.0 

17.0 

3.0 

3.5 

Aiatae  virginoparae 

2.6 

1.0 

3.8 

4.7 

1.8 

Oviparous  females 

1.4 

5.2 

Apterae  males 

0.8 

Aiatae  males 

0.2 

0.5 

Stillberg  (NE  slope) 

Eggs 

6.5 

2.0 

8.0 

Larvae 

1.0 

5.9 

7.7 

32.0 

40.0  89.0 

118.0 

26.0 

40.0 

58.0 

8.6 

12.0 

Fundatrices 

2.3 

0.8 

Apterae  virginoparae 

1.8  10.0 

18.0 

2.3 

7.1 

9.4 

3.8 

9.2 

Aiatae  virginoparae 

0.9 

1.0 

1.3 

3.0 

2.6 

2.8 

Oviparous  females 

0.8 

2.0 

Apterae  males 

0.7 

1.5 

Aiatae  males 

Stillberg  (N  slope) 

Eggs 

Larvae 

5.8 

43.0 

1.0 

11.0 

18.0 

Fundatrices 

Apterae  virginoparae 

6.2 

1.0 

2.0 

4.8 

Aiatae  virginoparae 

1.0 

3.4 

1.6 

1.2 

Oviparous  females 
Apterae  males 
Aiatae  males 


115 


SEASONAL  MOVEMENTS 

The  most  important  factors  affecting  aphid  feeding  and 
nutrition  are  the  change,  in  time,  of  growth  and  develop- 
ment of  the  host  plant  (Klungauf  1987).  In  conjmiction 
with  this  change,  seasonal  movements  of  colonies  were 
observed. 

The  fundatrix  generation  moves  from  the  needles  where 
overwintering  eggs  were  laid.  Feeding  sites  of  fimda- 
trices,  and  later  generations  of  virginoparae,  in  the  course 
of  Jime  and  first  half  of  July  were  on  last  year's  shoots. 
During  July,  colonies  moved  to  twigs  and  branches  in 
lower  parts  of  the  crown  (in  extreme  cases  some  of  the 
colonies  were  fovmd  on  stems).  This  was  just  after  the 
peak  of  the  annual  increment  of  the  shoots,  when  they 
start  to  become  lignified.  It  seems  to  be  related  to  de- 
scending sap  movement.  Conifer  lachnids,  as  phloem 
feeders,  exploit  food  sources  in  lower  parts  of  the  plant 
that  become  available  as  the  season  progresses. 

The  alatoid  larvae  of  alatae  virginoparae  usually  move 
from  the  colony  to  the  tops  of  the  shoots  before  their  final 
moult.  Here  they  moult  into  alatae  adults  and  fly  off  to 
colonize  similar  positions  on  other  plants. 

Simultaneously  with  the  occurrence  of  sexuales,  the 
colonies  moved  ascendentally.  At  the  end  of  the  season 
the  majority  of  the  individuals  were  on  last  year's  and 
current  season's  shoots.  From  this  position  oviparous 
females  moved  to  lay  eggs  on  needles. 

Additional  trials  were  carried  out  with  colonies  isolated 
by  polytene  covers,  which  prevented  migration  from  top  to 
bottom  parts  of  the  host  plant.  In  the  beginning,  the  den- 
sity of  isolated  colonies  was  much  higher  than  the  density 
of  nearby  nonisolated  colonies.  However,  early  occurrence 
of  alatae  forms  indicated  an  attempt  to  change  the  colo- 
nies' position. 

Later,  complete  starvation  of  the  population  was  caused 
by  restriction  of  movement.  It  was  concluded  that  the 
life  cycle  of  Swiss  stone  pine  aphid  is  very  much  synchro- 
nized with  the  growth  and  development  characteristics  of 
their  host  plant,  and  in  conjunction  with  this,  quality  of 
nutrition. 

PEST  DAMAGES 

Keen  (1938)  reported  that  several  species  of  lachnids 
cause  considerable  injury  to  conifers  in  North  America. 
Fumiss  and  Carolin  (1977)  described  visible  signs  of  in- 
jury, such  as  yellowing  of  needles  or  reduction  of  growth 
of  young  trees.  According  to  Seithner  (1936),  the  damage 
to  Swiss  stone  pine  is  quite  iinimportant,  but  he  believed 
that  the  trees  could  be  weakened  by  the  sucking  and  that 
secondary  pests  (for  example,  Pissodes  and  Pityophthorus) 
could  easily  attack  them. 

We  found  no  direct  visible  signs  of  damage  by  Swiss 
stone  pine  aphid  except  for  some  discolorations  on  feeding 
sites.  Our  observations  show  that  retardation  of  the 
growth  increment  by  Swiss  stone  pine  aphid,  if  any,  is 
much  less  significant  than  retardation  due  to  other  envi- 
ronmental factors. 

Indirectly,  Swiss  stone  pine  aphid's  honeydew  excre- 
tion, in  the  absence  of  heavy  rain,  accumulates  on  the 
surface  of  the  shoots  and  forms  an  ideal  substrate  for 


saprophytic  sooty  mould  fungi  to  develop.  A  black  cover- 
ing of  hyphae  and  spores  can  reduce  the  quality  of  trees 
that  are  grown  in  parks,  but  for  the  Stillberg  and 
Lucksalp  research  areas  it  is  not  so  important.  On  the 
other  hand,  honey  derived  from  honeydew,  particularly 
from  conifer  lachnids,  is  highly  valued  and  of  considerable 
economic  significance.  Fossel  (1971)  found  that  the  Swiss 
stone  pine  aphid  produced  copious  honeydew  in  the  movm- 
tainous  districts  of  Austria. 

CONCLUSIONS 

Information  gathered  during  this  study  has  allowed  the 
following  conclusions  to  be  drawn: 

•  Swiss  stone  pine  aphid  was  the  most  common  insect 
pest  on  Swiss  stone  pine  used  for  afforestation  in  the  re- 
search areas  of  the  Dischma  Valley. 

•  The  morphological  characteristics  of  specimens  and 
the  chronological  series  of  morphs  found  on  Stillberg  and 
Lucksalp  are  similar  to  the  conventional  descriptions. 
However,  for  some  site  aspects  of  the  Dischma  Valley  the 
life  cycle  of  Swiss  stone  pine  aphid  was  incomplete.  The 
life  cycle  abnormality  was  observed  on  the  north  slopes  of 
Stillberg. 

•  The  order  of  appearance  of  particular  insect  forms, 
structure,  and  number  of  individuals  in  colonies  also  were 
strongly  related  to  site  aspects.  This  is  particiilarly  true 
for  the  temperature  effect  that  caused  later  development 
and  smaller  sizes  of  colonies  on  the  north  slopes. 

•  Swiss  stone  pine  aphid  was  observed  to  cause  only 
unimportant  discolorations  on  the  feeding  sites.  The 
damage  to  the  trees  is  much  less  significant  than  that  pro- 
duced by  the  normally  harsh  environmental  conditions  of 
the  upper  timberline  area. 

When  looked  at  in  isolation,  the  damage  produced  by 
the  pest  observed  in  this  study  cannot  be  considered  seri- 
ous. However,  it  is  only  one  link  in  a  chain  of  factors  det- 
rimental to  the  growth  of  afforestation  plants.  How  the 
insect  interacts  with  other  factors  harmful  to  tree  growth 
(for  example,  fungi)  is  as  yet  unknown,  and  therefore  the 
damage  caused  by  colonies  of  Swiss  stone  pine  aphid  can- 
not be  accurately  assessed. 

Until  it  is  known  how  detrimental  the  aphid  is,  the  im- 
portance of  control  measures  or  predators  and  parasites 
also  cannot  be  estimated. 

Clearly,  6  months  investigation  is  too  short  for  con- 
clusive results.  However,  the  information  gained  has 
provided  an  important  first  step  in  imderstanding  the 
relationship  between  the  insect  pest,  Swiss  stone  pine 
aphid,  and  the  host  plant,  Swiss  stone  pine,  in  high- 
altitude  afforestation. 

REFERENCES 

Alikhan,  M.  A.  1960.  The  experimental  study  of  chemotac- 
tic  basis  of  host  specificity  in  a  phytophagous  insect. 
Aphis  fabae  Scop.  (Aphididae:  Homoptera).  Annales 
Univ.  Mariae  Curie  -  Sklodowska,  Sectio  C.  15:  117-158. 

Carter,  C.  I.;  Maslen,  N.  R.  1982.  Conifer  lachnids  in 
Britain.  Alice  Holt  Lodge,  Farnham,  Surrey,  Engleind: 
Forestry  Commission  Research  Station. 


116 


Chapman,  R.  F.;  Bemays,  E.  A.;  Simpson,  S.  J.  1981. 
Attraction  and  repulsion  of  the  aphid,  Cavariella 
aegopodii,  by  plant  odors.  Journal  of  Chemical  Ecology. 
7:  881-888. 

Dixon,  A.  F.  G.  1987.  Parthenogenetic  reproduction  and 
the  rate  of  increase  in  aphids.  In:  Aphids:  their  biology, 
natural  enemies,  and  control.  2A:  269-287. 

Eastop,  V.  F.  1972.  A  taxonomic  review  of  the  species 
of  Cinara  Curtis  occurring  in  Britain  (Hemiptera: 
Aphididae).  Bulletin  of  the  British  Museimi,  Natiiral 
History  (Entomology).  27:  103-186. 

Fossel,  A.  1971.  New  observation  on  Cinara  cembrae 
(Homoptera,  Lachnidae).  Annales  Zoologici.  28:  353-365. 

Fumiss,  R.  L.;  Carolin,  V.  M.  1977.  Western  forest  in- 
sects. Misc.  Publ.  1339.  Washington,  DC:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service.  654  p. 

Hille  Ris  Lambers,  D.  1955.  Hemiptera  2.  Aphididae. 
Zoology  of  Iceland.  3(52a):  1-29. 

Keen,  F.  P.  1938.  Insect  enemies  of  western  forests.  Misc. 
Publ.  273.  Washington,  DC:  U.S.  Department  of 
Agriculture. 

Klingauf,  F.  A.  1987.  Feeding,  adaptation  and  excretion. 
In:  Aphids:  their  biology,  natural  enemies,  and  control. 
2A:  225-253. 


Lees,  A.  D.  1961.  Clonal  polymorphism  in  aphids.  In:  In- 
sect polymorphism.  Sjmiposia  of  the  Royal  Entomologi- 
cal Society  of  London.  1:  68-79. 

Pintera,  A.  1966.  Revision  of  the  genus  Cinara  Curt. 
(Aphidodea,  Lachnidae)  in  Middle  Europe.  Acta  ent. 
bohemoslov.  63:  281-321. 

Pospisil,  J.  1976.  Sensory,  olfactory  and  visual  orientation 
of  insects  of  the  tropical  region  III.  Visual  and  olfactory 
orientation  of  alate  forms  of  the  aphid,  Aphis  gossypii 
Glover,  in  Cuba.  Academia  de  Ciencias  de  Cuba,  Serie 
Biologica.  66:  17-23. 

Schonenberger,  W.  1984.  Performance  of  a  high  altitude 
afforestation  under  various  site  conditions.  Proc.  3d 
lUFRO  Workshop:  233-240. 

Seither,  M.  1936.  Lachnus  cembrae  an.  sp.,  die  Zirben- 
blattlaus.  Centralbl.  ges.  Forstwes.  62:  33-49. 

Stroyan,  H.  L.  G.  1955.  Recent  additions  to  the  British 
aphid  faima.  Part  II.  Transactions  of  the  Royal  Entomo- 
logical Society  of  London.  106:  283-340. 

Stroyan,  H.  L.  G.  1960.  Three  new  subspecies  of  aphids 
from  Iceland  (Hem.,  Hom.).  Entomologiske  Meddelelser. 
29:  250-265. 


117 


GENETIC  CONSEQUENCES  AND 
RESEARCH  CHALLENGES  OF  BLISTER 
RUST  IN  WHITEBARK  PINE  FORESTS 

Raymond  J.  Ho^ 
Susan  K.  Hagle 
Richard  G.  Krebill 


Abstract — Susceptibility  of  whitebark  pine  {Pinus  albicaulis)  to 
blister  rust  (caused  by  Cronartium  ribicola)  is  reviewed.  Prog- 
ress is  reported  on  studies  that  assess  the  level  of  susceptibility 
over  its  entire  range  and  the  existence  of  resistance  in  various 
stands.  Two  breeding  approaches  are  discussed:  (1)  the  tradi- 
tional, where  trees  are  selected,  tested,  then  established  in  seed 
orchards;  (2)  a  natural  approach  that  aids  natural  processes  to 
establish  "a  natural  selection  stand." 


In  1910,  eastern  white  pine  (Pinus  strobus)  seedlings 
that  had  been  grown  in  Europe  were  planted  in  the  area 
of  Point  Grey,  BC,  Canada.  Some  of  these  seedlings  were 
infected  with  white  pine  blister  rust  (caused  by  the  fungus 
Cronartium  ribicola).  The  disease  was  finally  noticed  in 
the  fall  of  1921,  when  it  was  observed  on  endemic  western 
white  pine  (Pinus  monticola).  The  destructive  nature  of 
this  disease  in  North  American  white  pines  had  already 
been  documented  in  Europe  (Spaulding  1911). 

What  followed  was  a  frantic  rush  to  stop  blister  rust's 
spread  by  destroying  infected  trees,  eradicating  currants 
and  gooseberries  (genus  Ribes)  that  are  alternate  hosts  for 
the  disease,  and  using  several  promising  fungicides.  But 
this  work  failed,  and  the  fungus  kept  spreading.  By  about 
1960,  the  fungus  had  spread  throughout  most  the  range 
of  whitebark  pine  (Pinus  albicaulis)  (fig.  1),  which  also  in- 
cludes most  of  the  range  of  western  white  pine  and  sugar 
pine  (Pinus  lambertiana).  Examples  of  blister  rust  on 
whitebark  pine  are  illustrated  in  figure  2. 

The  first  infected  whitebark  pine  recorded  was  in 
the  arboretum  of  the  University  of  British  Columbia, 
Vancouver,  BC,  in  1922  (Bedwell  and  Childs  1943).  The 
first  discovery  of  blister  rust  on  native  whitebark  pine 
was  in  1926  on  the  Birkenhead  River  in  the  Coast  Range 
of  British  Columbia,  Canada  (Childs  and  others  1938). 
Intense  infection  and  mortality  by  blister  rust  was  not  far 
behind  (Bedwell  and  Childs  1943). 

The  degree  of  infection  was  100  percent  in  many  stands 
in  the  northern  areas,  and  decreased  to  the  south  where 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Raymond  J.  HofF  is  Research  Plant  Geneticist,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Moscow,  ID  83843, 
USA.  Susan  K.  Hagle  is  Pathologist,  Northern  Region,  Forest  Service,  U.S. 
Department  of  Agriculture,  Missoula,  MT  59807,  USA.  Richard  G.  Krebill 
is  Assistant  Station  Director  and  Pathologist,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Ogden,  UT  84401, 
U.S.A. 


whitebark  pine  occurs  in  higher  and  drier  sites.  White- 
bark pines  grovnng  below  about  45°  N.  lat.  in  Idaho  and 
Montana  have  much  less  rust  than  whitebark  of  higher 
latitudes.  However,  in  the  Cascade-Sierra  Nevada  Moun- 
tains the  rust  has  spread  to  about  lat.  36°  15'  N.,  where  it 


Pinus  albicaulis 

BH  malor  subalpine  component 


Figure  1 — Range  map  of  whitebark  pine  showing 
areas  of  high,  moderate,  and  little  or  no  blister  rust 
incidence. 


118 


has  been  especially  devastating  to  western  white  pine 
and  sugar  pine.  Whitebark  pine  too  has  been  heavily 
impacted  in  most  of  these  mountain  ranges.  But  no 
infection  was  found  on  whitebark  pine  at  Tioga  Pass, 
Yosemite  National  Park,  at  about  39°  N.  lat.  in  1992. 

The  relative  susceptibility  of  the  white  pines  to  blister 
rust  has  been  tested  by  several  workers.  Bingham 
(1972b)  siimmarized  these  reports  (table  1).  All  reports 
agreed  that  whitebark  pine  was  most  susceptible  of  the 
tested  five-needled  pines.  Fiu-ther  work  was  done  by 
Bedwell  and  Childs  (1943)  to  compare  the  relative  suscep- 
tibility of  whitebark  versus  western  white  pine.  These 
two  pines  grow  adjacent  for  much  of  their  ranges,  with 
whitebark  pine  at  the  high-elevation  sites  and  western 
white  pine  lower.  They  often  are  foimd  on  the  same  site 
in  the  overlap  zone.  Childs  and  Bedwell  (1948)  found  that 
susceptibility  (in  terms  of  nvunbers  of  cankers  and  speed 
of  tree  mortality)  of  whitebark  pine  was  several  times 
that  of  western  white  pine.  They  concluded  that  this  dif- 
ference was  due  in  part  to  (1)  a  higher  susceptibility  of 
current  year  whitebark  pine  needles  and  (2)  longer  needle 
retention  of  whitebark  pine  (5.3  years  for  whitebark  and 
3.8  years  for  western  white  pine). 


Table  1 — Comparison  of  susceptibility  in  white  pines  to  blister  rust 
by  Bingham  1972a  or  b  and  a  reshuffle  by  Hoff  and  others 
1980 


Species         Bingham  1972a  or  b      Hoff  and  others  1980 


Pinus  armandii 

1 

4 

P.  cembra 

1 

3 

P.  aristata 

1 

P.  wallichiana 

2 

6 

P.  koraiensis 

2 

1 

P.  peuce 

3 

5 

P.  sibirica 

4 

2 

P.  parviflora 

5 

3 

P.  strobiformis 

6 

8 

P.  strobus 

7 

11 

P.  flexilis 

8 

10 

P.  monticola 

9 

4 

P.  lambertiana 

10 

9 

P.  albicaulis 

11 

7 

119 


Bingham  (1983)  found  one  western  white  pine  in  10,000 
that  was  free  of  bhster  rust.  The  possibihty  of  finding 
disease-free  whitebark  pine  trees  seemed  pretty  sHm. 

Nonetheless,  at  sites  in  northern  Idaho  and  western 
Montana  where  mortahty  is  over  90  percent  due  to  blister 
rust,  we  have  found  a  few  individuals  with  no  or  just  a 
few  cankers.  Though  they  certainly  are  rare  (fig.  3).  This 
gives  us  hope  that  there  are  resistant  trees  that  can  be 
used  to  reestablish  populations  in  these  areas. 

Hofif  and  others  (1980)  presented  data  indicating  that 
three  phenotypically  resistant  whitebark  pine  trees  in- 
cluded in  a  species  trial  did  indeed  have  resistance  to  blis- 
ter rust.  In  this  same  study,  resistant  collections  of  west- 
em  white  pine  and  sugar  pine  were  also  included.  This 
resulted  in  a  reshuffle  of  the  white  pines  for  susceptibility 
to  blister  rust  (table  1).  In  this  comparison,  whitebark 
pine  moved  up  along  with  resistant  western  white  pine 
and  resistant  sugar  pine.  However,  these  same  species 
were  tested  in  France  by  Delatour  and  Birot  (1982),  in 
Japan  by  Yokota  (1983),  and  in  Germany  by  Stephan 
(1985),  and  the  resistance  level  of  the  collections  of  white- 
bark, western  white,  and  sugar  pine  was  much  lower 
(tables  2,  3,  4).  The  only  explanation  is  that  there  are 
different  races  of  blister  rust  at  these  other  test  sites. 

In  spite  of  this  bad  news,  our  plans  are  to  proceed  to  de- 
velop resistance  in  whitebark  pine,  including  as  many 


Figure  3 — Blister  rust-resistant  candidate  trees  at  Gisborne  Mountain  (A) 

120 


genes  for  resistance  as  possible.  If  new  races  of  the  fungus 
appear,  the  new  resistant  varieties  should  have  adequate 
flexibility  to  exhibit  resistance  to  the  new  races.  This 
seems  justified,  since  high  levels  of  resistance  in  western 
white  pine  and  sugar  pine  are  still  being  observed  in  blister 
rust  resistance  tests,  even  though  several  races  of  the  fun- 
gus are  evident. 

ACTIVE  WORK  AND  PRELIMINARY 
PROGRESS 

Blister  Rust  Surveys  of  Stands,  Clearcuts,  and 
Bums — Onr  objectives  are  to  determine  the  level  of  blis- 
ter rust  damage  over  the  reuige  of  whitebark  pine  and  to 
locate  phenotypically  resistant  trees.  As  mentioned,  the 
degree  of  damage  is  associated  with  latitude.  Phenotypi- 
cally resistant  trees  have  been  located  in  most  stands 
where  high  mortality  to  blister  rust  has  occurred.  We 
also  observed  a  relatively  large  amount  of  natural  regen- 
eration occurring  in  clearcuts  and  bums  adjacent  to 
stands  of  whitebark  pine  that  have  had  high  rust-caused 
mortality.  Table  5  shows  the  number  of  trees  per  acre 
and  the  degree  of  infection  by  blister  rust  in  four  such 
stands.  Natural  regeneration  of  whitebark  pine  occurs 
from  unused  seed  caches  of  Clark's  nutcracker.  The 


Burl^e's  Peak  (B). 


Table  2 — Percent  of  rust-infected  seedlings  in  bulk  collections  of 
white  pine  species  (from  Delatour  and  Birot  1 982) 


Geographic 

Number 

Percent 

group 

Species 

of  lots 

infected 

European 

P.  cembra 

12 

0 

P.  peuce 

7 

0-20 

Asian 

P.  sibirica 

2 

0 

P.  parviflora 

3 

0 

P.  koraiensis 

7 

0-9 

P.  wallichiana 

10 

7-67 

P.  armandii 

1 

58 

American 

r.  ailolala 

1 

o 
c. 

P.  albicaulis 

5 

40-100 

P.  flexilis 

17 

0-100 

P.  strobiformis 

9 

40-100 

P.  lambertiana 

5 

64-96 

P.  strobus 

24 

88-100 

P.  monticola 

5 

94-100 

abundance  of  regeneration  was  surprising.  It  immedi- 
ately broxight  up  questions:  At  what  level  of  mortality  will 
there  not  be  enough  seed  available  for  caching  by  the  nut- 
cracker? If  the  seed  did  not  come  from  the  adjacent  high- 
mortality  stand,  where  did  it  come  from? 

Artificial  Inoculation  With  Blister  Rust — The  objec- 
tive is  to  determine  the  level  and  variety  of  mechanisms 
that  impart  resistance  in  phenotypically  resistant  trees 
and  in  stands  that  have  had  high  mortality  by  blister 

Table  3 — Results  of  inoculation  in  France  and  Idaho. 

Percent  of  rust  infection  on  "resistant"  crosses 
of  P.  lambertiana  and  P.  monticola  from 
inoculation  in  France  or  Idaho  (Delatour  and 
Birot  1 982) 

Seed  lot  No.  France 
and  cross  mean  Idaho 


 Percent 

P.  lambertiana 


41  CF^ 

99 

55 

41  BF, 

98 

57 

41  AF, 

52 

25 

P.  monticola 

43  BF, 

96 

40 

43  EF, 

93 

30 

43  CF, 

76 

40 

43  A  F, 

32 

30 

43  FF, 

73 

65 

43  DF, 

35 

45 

47  A  F, 

75 

65 

46HF2 

63 

35 

46EF2 

56 

40 

46BF2 

55 

45 

46FF2 

44 

35 

46G  F2 

40 

30 

46DF2 

38 

20 

46  AFj 

37 

30 

46CF2 

24 

15 

Table  4 — Comparison  of  blister  rust  susceptibility  of  some  white 
pines  from  Stephan  (1985) 


North  America 

Percent 

Eurasia 

Percent 

species 

infection 

species 

infection 

P.  albicaulis 

97 

P.  armandii 

0 

P.  aristata 

66 

P.  cembra 

0 

P.  balfouriana 

90 

P.  koraiensis 

23 

P.  flexilis 

98 

P.  morrisonicola 

40 

P.  lambertiana 

^97  (76) 

P.  parviflora 

22 

P.  monticola 

99 

P.  peuce 

22 

P.  strobiformis 

88 

P.  pumila 

0 

P.  strobus 

100 

P.  sibirica 

17 

P.  wallichiana 

40 

'Two  seedlots  were  tested;  one  was  not  as  susceptible  as  the  other. 


rust.  A  trial  test  was  inoculated  in  August  1991  using 
methods  outlined  by  Bingham  (1972a).  This  test  included 
about  1,000  seedlings  from  36  families  from  10  populations. 
The  inoculation  was  light;  only  26  percent  of  the  seedlings 
developed  needle  spots.  However,  there  was  considerable 
variation  among  populations  (table  6).  Even  with  the  light 
inoculation,  it  was  encouraging  to  observe  several  seedlings 
that  had  what  we  call  "premature  needle  shed"  (McDonald 
and  Hoff  1970).  This  is  a  resistance  mechanism  in  which 
needles  with  needle  spots  are  shed  from  9  to  12  months 
after  inoculation  and  before  the  rust  fungus  has  invaded 
the  stem.  This  type  of  resistance  has  been  observed  in 
several  white  pine  species  (Hoff  and  others  1980).  The 
seedlings  were  reinoculated  in  August  1992  and  as  of 
November  1,  1992,  87  percent  had  needle  spots. 

Adaptive  Variation  of  Whitebark  Pine — How  far 

can  we  transfer  seed  £ind  not  cause  maladaptation  in  the 
seedlings?  The  range  of  whitebark  pine  extends  nearly 
19  degrees  of  latitude  and  20  degrees  of  longitude,  and 
from  1,524  to  3,354  m.  In  addition,  the  species  grows  in 
various  ecological  situations.  It  can  fill  the  role  of  a  pio- 
neer and  a  serai.  It  can  exist  for  long  periods  as  a  domi- 
nant on  harsh  sites  and  even  in  some  stands  with  the  cli- 
max species,  subalpine  fir.  Whitebark  pine  is  found  in 
106  of  225  habitat/phase  types  in  eastern  Idaho-western 
Wyoming,  central  Idaho,  and  Montana  (Pfister  and  others 
1977;  Steele  and  others  1981,  1983). 

However,  whitebark  pine  is  relatively  intolerant  to 
shade  and  highly  sensitive  to  competition  (Arno  and 
Weaver  1990).  Therefore  it  is  frequently  restricted  to 
tougher  sites  that  prevent  or  restrict  the  growth  of  other 


Table  5 — Number  and  infection  of  whitebark  pine  seedlings  in  four 
clearcuts 


Name  of  Seedlings  Blister  rust 

clearcut  per  hectare  infection 


Percent 

Vermillion  Pass 

445 

55 

Upper  Coal  Creek 

1,544 

4 

Upper  Big  Creek 

884 

33 

Divide  Mountain 

3,929 

52 

121 


Table  6 — Variation  in  various  traits  of  whitebark  pine  for  1 0  provenances 


Preliminary  population  comparisons 


Third-year 

riea 

hXi 

blister 

^Ji*  Aif  An  Q  n  A 

1  a* 

Lai. 

Long. 

Elevation 

height 

stem 

DUO 

rust 

m 

cm 

—  Percent-  - 

Cooper  Pass 

47-32N 

115-44W 

1,494 

17.1 

22 

9 

22 

Gisborne 

48-21 N 

116-44W 

1,692 

15.0 

10 

4 

24 

Freezeout 

47-01 N 

116-02W 

1,707 

10.6 

3 

0 

5 

Lunch  Peak 

48-22N 

116-22W 

1,982 

16.9 

10 

9 

33 

Brundage  Mountain 

45-01 N 

116-07W 

2,195 

16.2 

17 

8 

24 

Seven  Devils 

45-21 N 

11 6-31 W 

2,302 

12.8 

6 

3 

26 

Saddle  Mountain 

45-42N 

113-59W 

2,378 

18.1 

24 

2 

46 

Porphyry  Peak 

46-49N 

110-44W 

2,509 

12.0 

0 

13 

33 

Boulder  Peak 

41-35N 

1 23-05 W 

2,523 

10.7 

0 

0 

Palmer  Mountain 

45-04N 

110-35W 

2,652 

16.3 

0 

22 

41 

Mean 

14.6 

9 

7 

25 

species.  Because  of  the  combination  of  several  severe  en- 
vironmental factors,  such  as  short  growing  season,  cold 
air,  and  snow  blast  (McCaughey  and  Schmidt  1990), 
whitebark  pine  forms  krummholz  stands  of  shrublike 
trees  at  or  near  timberline.  At  lower  elevations,  where 
growing  conditions  are  not  quite  so  severe,  whitebark  pine 
grows  in  nearly  pure  stands  of  trees  30  to  90  feet  tall 
(McCaughey  and  Schmidt  1990).  When  environmental 
conditions  are  even  less  severe,  whitebark  pine  is  associ- 
ated with  other  tree  species  and  takes  on  a  form  much 
like  lodgepole  pine.  When  growdng  conditions  are  ideal, 
whitebark  pine  seedlings  will  grow  fairly  rapidly  (fig.  4). 

With  such  a  wide-ranging  and  ecologically  diverse  spe- 
cies one  would  expect  several  geographic  races.  Since 
nutcrackers  normally  cache  the  seed  at  least  several  him- 
dred  meters  and  in  many  cases  several  kilometers  from 
its  source,  it  is  also  easy  to  specvilate  that  there  are  few 
races  because  genes  from  several  diverse  populations  are 
being  continually  mixed.  Future  tests  should  clarify  this 
puzzle. 

Seed  from  about  40  stands  have  been  collected.  Our 
first  area  of  study  will  be  Idaho  and  Montana.  The  test 
will  be  established  with  seed  from  about  100  populations. 
A  preliminary  test  has  revealed  some  interesting  data 
(table  6).  There  is  considerable  variation  among  the  popu- 
lations in  3-year-old  total  height  (fig.  4).  The  tallest  seed- 
lings were  from  the  Saddle  Mountain  stand — a  population 
from  high  elevation.  There  is  also  high  variation  among 
families  within  stands  (table  7).  Also,  characteristics  such 
as  red  coloration  of  the  succulent  stem  and  the  prolifera- 
tion of  lateral  buds  vary  with  population  (fig.  5).  So  far, 
none  of  this  variation  can  be  associated  with  geographic 
position  or  elevation. 

Inbreeding — In  many  stands,  only  a  few  individuals 
remain.  This  results  in  much  higher  opportimity  for  self- 
fertilization  (inbreeding).  At  what  level  will  inbreeding 
adversely  affect  growth  and  survival  of  the  seedlings? 
A  study  has  been  initiated  to  compare  growth  and  other 
traits  between  a  highly  decimated  population  and  a  stand 
with  very  low  mortality.  Trees  at  the  two  sites  will  be  in- 
tercrossed, crossed  with  pollen  from  the  same  site,  and 
compared  v^dth  wind-pollinated  seed. 


BREEDING  PLANS  TO  DEVELOP 
RESISTANCE 

Our  surveys  have  shown  that  there  are  many  pheno- 
typically  resistant  whitebark  pine  trees.  A  traditional 
breeding  approach  could  be  used  to  develop  a  new  blister 
rust-resistant  variety  of  whitebark  pine  from  these  trees. 
Three  options  in  this  approach  will  be  discussed.  How- 
ever, because  of  the  high  level  of  natural  regeneration  of 
whitebark  pine  in  clearcuts  and  burns  adjacent  to  high- 
mortality  stands,  another  approach  appears  more  attrac- 
tive. This  is  a  more  natural  method  using  mass  selection 
as  the  genetic  selection  system,  but  within  a  natural  set- 
ting and  with  natural  selection  processes. 

The  Traditional  Approach  (fig.  6) — Steps  to  success 
are: 

1.  Locate  blister  rust-resistant  phenotypes  of  white- 
bark pine.  Most  stands  have  a  few  candidate  trees,  but 
a  high  number  is  required  so  that  we  can  be  assured  of 


Figure  4 — Three-year-old  container-grown  white- 
bark pine. 


122 


Table  7 — Variation  among  whitebark  pine  families  within  the 
Gisborne  Stand 


'amily 

Third-year 
neignt 

Red  stem 

Prolific 

DUO 

Diister 
rust 

cm 

-  -  Percent  — 

1 

18 

6 

0 

11 

2 

12 

11 

0 

22 

3 

16 

0 

0 

29 

4 

14 

12 

0 

0 

5 

15 

20 

5 

22 

6 

14 

7 

0 

14 

7 

17 

0 

0 

7 

8 

15 

4 

13 

33 

9 

18 

20 

4 

65 

10 

15 

9 

0 

35 

11 

12 

5 

9 

30 

Mean 

15 

9 

3 

24 

finding  several  genes  for  resistance.  Thus,  many  stands 
will  have  to  be  visited.  The  breeding  population  for  west- 
em  white  pine  is  over  3,000  phenotypically  resistant  trees 
and  is  considered  a  minimum  nimiber  of  parent  trees  to 
assure  maintenance  of  resistance  against  several  races  of 
the  rust  fungus. 

2.  Collect  wind-pollinated  seed  from  each  tree.  In  spe- 
cial cases,  where  there  are  just  one  or  two  trees,  artificial 
pollination  by  other  candidate  trees  may  be  advisable. 

3.  Sow  seed,  grow  seedlings,  and  inoculate  with  blister 
rust  in  the  fall  of  their  second  growth  period. 

4.  Data  over  the  next  4  years,  after  inoculation,  will 
reveal  the  most  resistant  parents  of  the  phenotypically 
resistant  candidate  trees,  along  with  their  resistant 
progeny. 

Options  for  developing  the  new  resistant  variety  are: 

1.  Selected  resistant  parents  can  be  grafted  into  a  seed 
orchard. 

2.  Resistant  seedlings  can  be  used  to  establish  a  seed 
orchard. 

3.  Surviving  seedlings  from  rust  tests  that  were  not 
used  in  the  seed  orchard  can  be  outplanted  in  a  natural 
site.  The  numbers  for  this  planting  can  probably  be  in- 
creased by  rooted  cuttings  of  the  surviving  seedlings. 

Timing:  The  first  five  steps  cotdd  be  done  in  about  10 
years.  Time  to  flowering  of  grafts  could  be  10  years,  but  it 
will  probably  take  from  40  to  50  years  for  their  progeny  to 
flower.  New  technology  may  shorten  this  time. 

Problems: 

1.  What  is  the  natural  level  of  resistance? 

2.  How  far  can  seed  be  transferred  without  serious 
maladaptation? 

Natural  Selection  Stand — In  many  localities  in  Idaho 
and  Montana  whitebark  pine  has  been  decimated  by  blis- 
ter rust,  leaving  just  a  few  individuals  per  hectare.  Many 
of  these  remnant  stands  are  adjacent  to  burns  or  clear- 
cuts.  In  at  least  several  dozen  cases  these  openings 


contain  high  nvunbers  of  whitebark  pine  seedlings  and 
saplings  (table  5).  With  careful  management  we  may  be 
able  to  use  these  newly  established  stands  for  our  piir- 
pose.  Only  three  major  steps  need  to  be  taken: 

1.  Select  the  areas  within  the  clearcuts  and  bxarns  that 
will  be  managed  for  whitebark  pine  (fig.  7A). 

2.  Clean  the  selected  area  of  competing  trees  and 
shrubs  (fig.  7B).  Whitebark  pine  is  a  poor  competitor.  In 
fact  it  probably  can  be  grown  at  much  lower  elevations  if 
the  site  is  kept  free  of  competition. 

3.  Let  nature  take  its  course  concerning  the  selection  of 
the  most  resistant  trees  (figs.  7C,  D). 

Problems: 

1.  What  is  the  natural  level  of  resistance?  We  need  to 
know  the  level  of  resistance  (1,  5,  or  10  percent)  in  stands 
to  choose  sites  that  have  a  good  chance  for  success.  If  only 
1  percent  of  the  seedlings  are  resistant  when  produced  by 
a  mature  stand  that  has  had  90  to  95  percent  mortality 
by  blister  rust,  a  candidate  area  woidd  have  to  have  an 
unusually  high  number  of  seedlings  to  end  up  with  a  rea- 
sonably stocked  stand.  If  10  to  20  percent  are  resistant, 
the  probability  of  producing  a  new  stand  would  be  consid- 
erably better. 


Figure  5 — Proliferated  lateral  buds  on  whitebark  pine. 


123 


Locate  Parents 
(Resistant  Phenotypes) 

Collect  Wind  or 
Cross-Pollinated  Seed 

Blister  Rust  Test 
1 0  Years 


Graft  Seed  Orchard 
Production 
1 0  Years 


Seedling  Seed  Orchard 
Production 
40  Years 


Outplant  Survivors 
From  Blister  Rust 
Test 


Stand  Establishment 
50  Years 


Stand  Establishment 
50  Years 


Stand  Establishment 
50  Years 


Total  Time 
70  Years 


Total  Time 
100  Years 


Total  Time 
60  Years 


Figure  6 — Options  for  traditional  breeding  methods  for  developing  new  varieties  of  blister  rust-resistant 
whitebark  pine,  including  completion  times. 


2.  How  much  inbreeding  is  occurring  in  the  decimated 
stands?  Severe  inbreeding  might  result  in  reduced 
growth  and  survivabihty. 

3.  What  is  the  minimimi  number  of  cones  £ind  seed 
needed  before  seed  caches  are  effective  for  regeneration? 
Some  sites  may  be  beyond  help. 

4.  When  these  sites  start  producing  resistant  seed  how 
far  can  the  seed  be  transferred? 

Timing  (fig.  8):  First-generation  blister  rust  selection 
has  already  been  made  on  parent  trees.  Seedling  estab- 
lishment has  already  occurred  on  dozens  of  sites.  Natursd 
selection  by  blister  rust  is  occmring.  In  40  to  50  years  es- 
tablished stands  on  preferred  sites  could  be  producing 
cones. 

However,  it  is  not  only  the  increase  of  resistance  genes 
that  is  important — although  this  is  a  key  to  the  species' 
future — it  is  important  to  maintain  as  much  of  the  natu- 
ral function  of  whitebark  pine  forests  as  possible.  In  addi- 
tion to  the  severe  genetic  impact  to  whitebark  pine  itself, 
the  critical  ecological  functions  of  this  species  have  been 
threatened  by  loss  of  integrity  of  whitebark  pine  forests. 
Thresholds  are  a  key  consideration  when  attempting  to 
restore  severely  altered  ecological  systems  (Friedel  1991; 
Laycock  1991).  The  ecological  thresholds  for  whitebark 
pine  forests  are  not  well  understood  and  with  the  loss  of 
most  of  the  original  structure  of  these  forests  much  of  the 
integrity  has  been  lost.  It  will  take  generations  to  repro- 
duce nearly  pure,  mature  whitebark  pine  forests.  In  the 
interim,  components  of  the  ecological  systems  most  af- 
fected by  the  loss  of  whitebark  pine  forests  may  be  irre- 
versibly altered. 

Time  is  of  the  essence,  but  all  is  not  lost.  Over  large  ar- 
eas the  first  generation  is  mostly  gone,  but  in  many  areas 


the  second  generation  is  there;  we  must  act  now  to  save  it. 
In  other  sireas  the  first  generation  is  largely  intact.  In 
these  areas  immediate  emphasis  needs  to  be  placed  on  re- 
taining or  restoring  the  ecological  function  of  whitebark 
pine.  Protective  measures,  such  as  pruning  and  Ribes 
population  manipidation  may  be  used. 

What  we  are  trying  to  say  in  this  section  is  that  there  is 
much  that  can  and  must  be  done  immediately  to  maintain 
and  restore  whitebark  pine  in  the  ecosystem. 

BLISTER  RUST  ON  WHITEBARK 
PINE,  A  GENETICS  PERSPECTIVE 

The  general  level  of  resistance  of  the  original  popula- 
tion (before  the  introduction  of  blister  rust)  is  less  than 
1  percent.  Therefore,  the  genetic  consequence  of  the  blis- 
ter rust  epidemic  for  most  of  the  range  of  whitebark  pine 
is  to  push  it  through  an  extremely  small  bottleneck.  In 
some  areas  where  the  number  of  surviving  trees  has  been 
reduced  to  only  a  few  over  several  hundred  hectares,  the 
bottleneck  may  result  in  entire  populations  being  lost. 
Figure  1  indicates  geographic  populations  of  whitebark 
pine  that  will  have  the  most  immediate  difficulty  (high  in- 
cidence area),  those  that  will  have  only  minor  problems 
(moderate  incidence  area),  and  those  that  are  not  likely  to 
have  a  problem  (little  or  no  incidence).  Fiulher  research 
clarifying  levels  of  natural  resistance  will  be  a  significant 
aid  to  predicting  the  adequacy  of  natural  selection  in  the 
various  areas.  On  some  sites,  time  alone  may  return 
whitebark  pine  to  its  former  levels;  on  other  sites,  white- 
bark pine  may  be  lost  if  not  aided  by  genetic  improvement 
and  silvicultural  programs. 

In  areas  where  there  are  many  surviving  trees  within  a 
very  reduced  popidation,  the  concern  will  be  the  damage 


124 


Whitebark  Pine  Competing  Species 

Figure  7 — Development  of  a  natural  selection  stand:  A,  A  clearcut  before  cleaning  containing  whitebark 
pine  and  competing  trees;  B,  clearcut  after  cleaning;  C,  clearcut  after  infection  by  blister  rust  and  selection 
of  most  resistant  trees;  D,  the  natural  selection  stand  or  seed  orchard  after  growth  of  selected  trees. 


Locate  Sites 
Remove  Competing 
Trees 


Stand  Establishment 
50  Years 


Some  stands  are 
already  20  years  old 


Total  Time 
50  Years  or  Less 


Figure  8 — Natural  selection  method  for  developing 
a  new  vahety  of  blister  rust-resistant  whitebark  pine, 
including  completion  times. 


done  by  inbreeding.  In  most  conifers  inbreeding  is  detri- 
men1:al.  Inbred  trees  generally  grow  slowly,  and  detrimen- 
tal genes  such  as  for  albinism,  short  leaves,  and  stunted 
seedlings  may  often  occur.  However,  whitebark  pine  is  of- 
ten transported  by  birds  to  areas  where  there  is  no  other 
whitebark  pine  (Linhart  and  Tomback  1985).  Because 
there  are  only  a  few  individuals  in  the  start-up  stand,  in- 
breeding would  be  high.  This  may  have  been  the  most 
common  way  of  past  population  expansion  and,  over  time, 
lethal  and  other  detrimental  genes  may  have  dropped  out 
of  the  population  leaving  a  species  that  is  not  adversely 
affected  by  inbreeding. 

Even  vTith  all  these  dire  genetic  consequences,  the  ge- 
netic opportimity  for  the  survival  of  whitebark  pine  in  the 
high-mortality  zone  appears  very  high,  especially  with  the 
intelHgent  help  of  humans.  There  are  many  phenotypi- 
cally  resistant  trees.  These  surely  can  be  used  as  a  base 


125 


for  new  varieties  of  whitebark  pine  resistant  to  blister 
rust.  But  most  exciting,  the  high  natural  regeneration  of 
whitebark  pine  stands  in  clearcuts  and  burns — even  in  ar- 
eas where  there  appear  to  be  insufficient  nimibers  of  par- 
ent trees — indicates  that  natural  processes  are  already 
producing  a  new  variety.  Further,  these  trees  are  grow- 
ing well;  there  do  not  appear  to  be  inbreeding  problems. 

CONCLUSIONS 

The  consequences  of  the  blister  rust  epidemic  have  to  be 
viewed  as  a  disaster  for  whitebark  pine,  and  a  setback  for 
Clark's  nutcrackers,  grizzly  bears,  and  other  components 
of  whitebark  pine  forest  ecosystems.  Nevertheless,  white- 
bark pine  appears  to  be  equipped  with  elements  that  will 
permit  survival.  Most  important  is  seed  caching  by  the 
nutcracker  even  when  there  is  just  a  small  excess.  Also, 
phenotypically  resistant  trees  seem  to  be  fairly  common. 
Considering  the  low  numbers  of  noncankered  western 
white  pine  trees  together  with  the  much  higher  suscepti- 
bility of  whitebark  pine,  we  wonder  that  there  are  any 
whitebark  pine  trees  at  all. 

Over  large  areas,  the  first  generation  of  whitebark  pine 
is  nearly  gone;  however,  the  second  generation  is  often 
present.  In  other  areas  the  first  generation  is  still  largely 
intact.  To  save  whitebark  pine  and  its  ecosystem  in  the 
most  sensitive  areas  it  is  essential  that  we  act  now.  Main- 
taining sufficient  ecosystem  function  for  whitebark  pine 
while  developing  greater  resistance  will  be  a  challenge  for 
land  managers. 

REFERENCES 

Arno,  S.  F.;  Weaver,  T.  1990.  Whitebark  pine  community 
types  and  their  patterns  on  the  landscape.  In:  Schmidt, 
W.  C;  McDonald,  K  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  97-105. 

Bedwell,  J.  L.;  Childs,  T.  W.  1943.  Susceptibility  of  white- 
bark pine  to  blister  rust  in  the  Pacific  Northwest.  Jour- 
nal of  Forestry.  41:  904-912. 

Bingham,  R.  T.  1972a.  Artificial  inoculation  of  large  num- 
ber of  Pinus  monticola  seedlings  with  Cronartium 
ribicola.  In:  Bingham,  R.  T.;  Hofi",  R.  J.;  McDonald, 
G.  I.,  eds.  Biology  of  rust  resistance  in  forest  trees. 
Misc.  Publ.  1221.  Washington,  DC:  U.S.  Department  of 
Agriculture:  357-372. 

Bingham,  R.  T.  1972b.  Taxonomy,  crossability,  and  rela- 
tive blister  rust  resistance  of  5-needled  white  pines.  In: 
Bingham,  R.  T.;  Hoff,  R.  J.;  McDonald,  G.  I.,  eds.  Biol- 
ogy of  rust  resistance  in  forest  trees.  Misc.  Publ.  1221. 
Washington,  DC:  U.S.  Department  of  Agric;ilture: 
271-280. 

Bingham,  R.  T.  1983.  Blister  rust  resistant  western  white 
pine  for  the  Inland  Empire:  the  story  of  the  first  25 
years  of  the  research  and  development  program.  Gen. 
Tech.  Rep.  INT-146.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermovmtain  Forest  and 
Range  Experiment  Station.  45  p. 


Childs,  T.  W.;  Bedwell,  J.  L.  1948.  Susceptibility  of  some 
white  pine  species  to  Cronartium  ribicola  in  the  Pacific 
Northwest.  Journal  of  Forestry.  46:  595-599. 

Childs,  T.  W.;  Bedwell,  J.  L.;  Englerth,  R.  H.  1938.  Bhster 
rust  infection  of  Pinus  albicaulis  in  the  Northwest. 
Plant  Disease  Reporter.  22:  139-140. 

Delatour,  C;  Birot,  Y.  1982.  The  international  lUFRO  ex- 
periment on  resistance  of  white  pines  to  blister  rust 
{Cronartium  ribicola):  the  French  trial.  In:  Heybroek, 
H.  M.;  Stephan,  B.  R.;  Weissenberg,  K.,  eds.  Resistance 
to  diseases  and  pests  in  forest  trees.  Wageningen: 
Pudoc:  412-414. 

Friedel,  M.  H.  1991.  Range  condition  and  the  concept  of 
thresholds:  a  viewpoint.  Journal  of  Range  Management. 
44:  422-426. 

Hoff,  R.  J.;  Bingham,  R.  T.;  McDonald,  G.  I.  1980.  Relative 
blister  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Laycock,  W.  A.  1991.  Stable  states  and  thresholds  of 
range  condition  on  North  American  rangelands:  a  view- 
point. Journal  of  Range  Management.  44:  427-433. 

Linhart,  Y.  B.;  Tomback,  D.  F.  1985.  Seed  dispersal  by 
nutcrackers  causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine  (Pinus  albicaulis  Engelm.).  In:  Schmidt, 
W.  C;  McDonald,  K  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  85-96. 

McDonald,  G.  I.;  Hoff,  R.  J.  1970.  Resistance  to  Cro- 
nartium ribicola  in  Pinus  monticola:  early  shedding  of 
infected  needles.  Res.  Note  INT-124.  Ogden,  UT:  U.S. 
Department  of  Agriculture,  Forest  Service,  Intermoun- 
tain  Forest  and  Range  Experiment  Station.  8  p. 

Pfister,  R.  D.;  Kovalchick,  R.  L.;  Arno,  S.;  Presby,  R.  1977. 
Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Forest  and  Range 
Experiment  Station.  174  p. 

Spaulding,  P.  1911.  The  blister  rust  of  white  pine.  Bull. 
206.  Washington,  DC:  U.S.  Department  of  Agriculture, 
Bureau  of  Plant  Industry.  88  p. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.  W.; 
Pfister,  R.  D.  1983.  Forest  habitat  t3rpes  of  eastern 
Idaho-western  Wyoming.  Gen.  Tech.  Rep.  INT-144. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Forest  and  Range  Experiment  Sta- 
tion. 122  p. 

Steele,  R.;  Pfister,  R.  D.;  Ryker,  R.  A.;  Kittams,  J.  A.  1981. 
Forest  habitat  types  of  central  Idaho.  Gen.  Tech.  Rep. 
INT-114.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  138  p. 

Stephan,  B.  R.  1985.  Resistance  of  five-needle  pines  to 
blister  rust.  Allgemeine  Forstzeitschrift.  No.  28: 
695-697. 

Yokota,  S.  1983.  Resistance  of  improved  Pinas  monticola 
and  some  other  white  pines  to  the  blister  rust  fungus, 
Cronartium  ribicola,  of  Hokkaido,  Japan.  European 
Journal  of  Forest  Pathology.  13:  389-402. 


126 


PERFORMANCE  OF  PINUS  CEMBRA, 
P.  PEUCE,  AND  P.  STROBIFORMIS 
WITHIN  AIR-POLLUTED  AREAS 

Karel  Kahak 


Abstract — ^An  inventory  of  the  surviving  species  within  the 
heavily  air-polluted  area  of  the  Ore  Mountains  showed  a  number 
of  species  of  the  genus  Pinus  without  symptoms  of  injury.  Three 
species  of  the  stone  pines  are  described  to  demonstrate  the  im- 
portance of  the  geological  past  of  the  species.  It  relates  to  the 
impact  of  the  catastrophes  and  volcanic  activities  and  sudden 
changes  of  climate  for  the  species'  preadaptation  to  the  toxic 
impact  of  a  changing  environment.  The  causes  of  the  resistance 
seemed  to  be  connected  with  the  long-term  migration  {Pinus 
cembra,  P.  peace)  and  hybrid  origin  (P.  peace,  P.  strobiformis). 
The  importance  of  the  life  history  of  the  species  is  unquestioned. 


During  the  years  1956  until  1966  the  author  (Kahak 
1971,  1988,  1991)  estabhshed  an  international  collection 
of  species  of  gentis  Pinus  in  Plzeh,  West  Czechslovakia, 
in  the  forest  locality  of  Sofronka  (latitude  50°  N.). 

Various  species  of  pines  are  represented  in  Sofronka  ei- 
ther by  patterns  of  populations  that  come  from  different 
localities  within  every  species  range  or  from  groves  of  in- 
dividual trees  from  species  that  originated  from  locations 
that  were  not  sufficiently  preadapted  for  the  local  condi- 
tions of  the  new  environment  (Kahak  1988).  A  research 
center  was  established  at  the  Arboretxim  Sofronka  in  1963 
to  investigate  evolution  of  the  species  included  within  the 
collection.  The  entire  range  of  provenance  experiments 
with  some  of  the  species  was  analyzed  with  respect  to  the 
evolution  of  their  distribution  area  during  postglacial 
time. 

The  air  pollution  caused  by  extensive  industrial  emis- 
sions resulted  in  a  mass  extinction  of  the  Norway  spruce 
forests  in  the  Ore  Mountains  in  the  vicinity  of  Plzeh.  In 
the  first  stage  of  the  extinction,  the  dying  spruce  stands 
were  cleared,  but  small  groves  of  individual  trees,  pre- 
dominantly the  local  and  exotic  species  of  pines,  survived. 
Therefore,  the  staff  of  the  research  center  at  Sofronka  was 
invited  to  join  the  search  for  an  ecologically  harmless 
method  of  reconstructing  the  extinct  forests. 

Some  species  of  white  pines  were  discovered  whose 
needles  appeared  to  be  unaffected  by  emissions.  These 
species  of  stone  pines  were:  Swiss  stone  pine  {Pinus 
cembra  L.),  Macedonian  pine  (Pinus  peuce  Griseb.),  and 
southwestern  white  pine  {Pinus  strobiformis  Engelm.). 
Their  unique  viability  had  helped  them  to  survive  under 
the  highly  unfavorable  conditions. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Karel  Ka&fik,  National  Park  of  Simiava,  Vimperk,  Czechoslovakia. 


SWISS  STONE  PINE 

1.  Occurs  as  individual  trees  in  the  entfre  emission 
area  between  the  town  of  Hora  Sv.  Sebestiana  and  Mount 
Klmovec  (1,210  m  above  sea  level  [a.s.l.]).  Some  trees, 
otherwise  injured  by  snow  and  icing,  show  no  injury  from 
emissions.  Many  of  them  bear  germinant  seeds  (Mottl 
and  Prudic  1982). 

2.  One  grove  of  fertile  trees  60  to  70  years  of  age  occurs 
in  the  cemetery  arboretiun  in  the  town  of  Vejprty.  Their 
dimensions  are  12  cm  diameter  at  1.3  m  and  9-11  m 
height.  They  bear  germinant  seeds. 

MACEDONIAN  PINE 

On  the  Saxon  side  of  the  mountains: 

1.  Forest  district  Steinbach,  Department  157a2,  eleva- 
tion 800  m  a.s.l.  A  group  of  17  trees,  age  of  55  years, 
fertile,  with  natural  regeneration.  Dimensions:  Diameter 
26-35  cm  at  1.3  m  and  15  m  height. 

2.  Forest  district  Schmalzgrube,  Department  291a3, 
elevation  900  m  a.s.l.  Twelve  trees  of  the  same  prove- 
nance and  age  as  the  latter  growing  on  a  peat  bog!  The 
trees  are  healthy  and  fertile.  Dimensions:  Diameter 
15-40  cm  at  1.3  m  and  11-14  m  height.  Tree  tnmks  have 
been  injured  by  the  red  deer  and  humans  (chmbing 
sptars). 

3.  Forest  district  Grumbach,  Department  331a4,  ele- 
vation 825  m  a.s.l.  Stand  area  is  0.75  ha  and  trees  are 
50  years  old  (1982)  and  grow  in  a  mixture  with  Pinus 
contorta  ssp.  latifolia  and  the  movmtain  variant  of  Pinus 
sylvestris  L.  Average  height  of  the  stand  is  18  m,  diam- 
eter is  21-35  cm  at  1.3  m.  Macedonian  pine  is  the  most 
vital  of  the  named  species. 

On  the  Czech  side  of  the  mountains: 

1.  Locality  Bludna,  Forest  Management  Horm'  Blatna, 
elevation  1,000  m  a.s.l.  Northern  slope  stand  of  approxi- 
mately 200  trees,  age  probably  70  years.  Origin  tmknown, 
but  probably  from  activities  of  the  Forest  Research  Insti- 
tute Mariabnmn  just  before  World  War  I.  Many  trees  were 
apparently  broken  at  the  height  of  5-7  m,  but  replacement 
crowns  are  now  reaching  heights  of  15  m.  Trees  are  rela- 
tively fertile  with  natural  regeneration  present. 

SOUTHWESTERN  WHITE  PINE 

1.  Locality:  Cemetery  arboretum  in  the  town  of  Vejprty 
with  a  heavy  emission  impact.  One  tree  only.  Diameter 
36  cm  at  1.3  m,  height  14  m,  and  age  50-60  years.  Tree  is 
fertile  with  germinant  seeds  in  the  cones  Ijdng  on  the 
moist  grass! 


127 


PROBABLE  CAUSES  OF  RESISTANCE 
Autochtonous  Species 

Ten  years  of  observations  and  experimental  activi- 
ties in  the  7-year  period  1982-89  give  us  the  following 
indications: 

Variability — ^Among  the  local  forest  tree  species,  there 
are  different  degrees  of  resistance  between  individual 
trees  within  the  stands.  Both  autochtonous  and  alloch- 
tonous  pine  species  had  been  gradually  reduced,  but  there 
are  some  examples  without  any  symptoms  of  injury.  Fir 
disappeared  on  the  Czech  side  of  the  mountain  range 
first,  but  drought  rather  than  emissions  may  have  been 
the  cause.  Norway  spruce  seems  to  be  extinct. 

Catastrophic  Selection — Norway  spruce  mostly  dis- 
appeared in  a  relatively  short  time — about  6  years.  How- 
ever, after  its  massive  extinction,  individually  resistant 
trees  of  this  species  were  found  within  the  extinct  forests. 
This  appearance  is  commonly  known  under  the  term 
"population  bottleneck"  (Dobzhanski  and  others  1976)  or 
"catastrophic  selection"  (Lewis  1962;  Raven  1964)  or  "une 
evolution  acceleree  (Bouvarel  1960). 

Founder  Effect — ^According  to  Mayr's  "founder-effect" 
(Mayr  1942,  1965,  1979)  after  the  strong  reduction  of 
Norway  spruce  the  reconstruction  of  its  genetic  architec- 
ture within  the  small  number  of  survivors  (founder  popu- 
lation) might  follow  (Kahak  1988).  This  might  give  rise 
to  a  new  species  population  endemic  to  the  environmental 
conditions  that  caused  the  latter  extinction. 

Allochtonous  Species 

Hybridization — The  Canadian  species  of  spruce  show 
a  high  degree  of  emission  resistance  compared  to  the  local 
Norway  spruce.  Their  life  history,  with  their  intricate 
genetic  relations,  may  help  explain  this  orientation 
(Heimburger  and  others  1983;  Morgenstern  and  Farrar 
1964).  Morphological  and  genetic  evidence  indicates  that 
black  spruce  is  a  relatively  young  species.  It  is  thought  to 
be  of  an  American  origin  resulting  from  the  hybridization 
of  a  proto- white  spruce  from  eastern  Asia  migrating  east- 
ward with  a  proto-red  spruce  of  North  American  origin 
migrating  westward.  Differences  in  morphological  char- 
acteristics between  black  spruce  and  red  spruce  in  the  di- 
rection of  white  spruce  would  indicate  such  hybrid  origin 
(Heimburger  1983). 

Directional  Selection 

The  most  relevant  information  came  from  all  the  for- 
eign species  formerly  introduced  to  this  mountain  range. 
An  excellent  example  is  the  life  history  of  lodgepole  pine 
as  described  by  Hansen  (1942,  1943).  This  species  came 
from  the  Cascade  Range  territories  of  North  America 
where  it  had  been  influenced  by  repeated  volcanic  activ- 
ity. Consequently,  many  generations  of  lodgepole  pine 
were  subjected  to  directional  selection  through  the  toxic 
air  and  soil,  and  lodgepole  pine  adapted  to  become  the 
founding  pioneer  stage  of  the  ecosystems'  evolution. 


Migration  Preadaptation 

The  other  very  important  evolutionary  factor  is  migra- 
tion. It  acts  on  the  development  of  ecological  potentisds 
as  a  consequence  of  the  tolerance  to  the  extreme  envi- 
ronmental conditions  (toxic  sites),  the  environmental 
changes,  and  all  the  preadaptations  that  could  be  ac- 
quired only  during  a  long  migration  route  such  as  those 
in  the  Tertiary — one  from  eastern  Asia  to  southern 
Europe  (Mirov  1967)  and  the  other  from  eastern  Asia 
to  Australia  (Mayr  1965). 

HISTORY  OF  SPECIES 

Mirov  (1967)  located  the  center  where  the  genus  Pinus 
originated  in  Eastern  Asia,  where  catastrophes  like  vol- 
canic activity  and  changes  in  climate  gave  all  the  species 
of  pines  a  specific  starting  point  to  their  principal 
adaptation. 

"Pines  however  possessed  two  characteristics  since  the 
Mezozoic  origin  of  the  genus:  their  xeromorphy  and  their 
ability  to  endure  direct  sunlight.  These  characteristics 
may  be  considered  as  generally  the  same  throughout  the 
whole  paleobotanical  history."  (Mirov  1967). 

Pinus  cembra 

"During  the  Ice  Age,  the  high  mountain  pines  of  the 
Alps  either  perished  or  were  preserved  in  the  lower  areas 
not  covered  with  ice.  Pinus  cembra  found  occasionally  in 
northern  Italy  is  the  relic  preserved  in  one  of  such  refugia 
(Emberger  1944).  With  subsequent  warming  up,  it  mi- 
grated again  to  the  high  elevations  of  the  central  Alps. 
There  is,  of  course,  a  possibility  of  the  existence  of  non- 
glaciated  refugia  even  in  the  central  parts  of  the  Alps. 
Pinus  cembra  might  have  survived  in  some  of  these  few 
areas  that  remained  free  from  the  ice."  (Mirov  1967). 

The  indications  are  that  P.  cembra  and  many  other  spe- 
cies of  the  eastern  Asia  flora  migrated  to  the  Alps  during 
the  Tertiary.  During  the  last  glaciation  they  occurred  in 
an  almost  continuous  belt  of  higher  elevations.  Conse- 
quently, because  of  the  enlarged  area  of  its  distribution, 
P.  cembra  developed  a  well-supported  degree  of  genetic 
variability.  This  species  of  stone  pine  survived  the  last 
glaciation,  and  within  its  refugias  there  was  probably  an 
increasing  frequency  of  autogamy  that  must  be  taken  into 
account  (Bannister  1965).  After  the  retreat  of  the  glacier, 
the  migrating  progenies  probably  intercrossed  at  their 
contact  zone  and  thus  acquired  their  increasing  variabil- 
ity as  well.  Together  with  the  very  short  period  of  the 
last  glaciation,  the  imdiminished  variability  still  main- 
tains itself  today.  Pinus  peuce  apparently  has  a  similar 
pattern  of  the  life  history. 

Pinus  peuce 

According  to  Mirov,  "There  are  indications  that  Medi- 
terranean pines  migrated  from  eastern  Asia  along  the 
mountain  ranges,  that  once  extended  north  of  and  parallel 
to  Himalaya."  (Mirov  1967). 

"Endemic  Pinus  peuce  is  very  local  in  Bulgaria,  Albania 
and  Yugoslavia.  It  is  a  haploxylon  pine  closely  related  to 


128 


Pinus  griffithii  of  the  Himalaya.  Before  the  great  migra- 
tion, Pinus  pence  has  been  crossed  with  other  haploxylon 
pines.  Judging  by  the  chemical  composition  of  its  tiirpen- 
tine  (Illoff  and  Mirov  1956)  it  has  a  close  affinity  to  east- 
em  Asiatic  and  western  American  pines"  (Mirov  1967). 

The  hybridization  should  be  a  source  of  variation  for  ad- 
aptation to  new  environments  (Lewontin  and  Birch  1966). 

The  whole  Mediterranean  region  experienced  many  tec- 
tonic changes  throughout  the  Tertiary.  In  the  Quarter- 
nary  there  occurred  intensive  faulting  and  sinking  of 
considerable  areas.  This  resiilted  in  the  formation  of  sep- 
arate moimtain  ranges  and  many  islands.  As  a  conse- 
quence, the  present  distribution  of  pines  in  the  Mediter- 
ranean region  is  sporadic  and  irregular  (Mirov  1967). 

Pinus  strohiformis 

Southwestern  white  pine  appear  to  have  morphological 
indications  of  a  hybrid  origin.  Critchfield  and  Little 
(1966)  said:  "This  species  forms  a  link  both  geographi- 
cally and  morphologically,  between  the  neighbors  Pinus 
flexilis  and  Pinus  ayacahuite.  Southwestern  pine  is  ap- 
parently separated  from  its  southern  relative  Pinus  aya- 
cahuite by  a  100  mile  gap  that  does  not  seem  to  corre- 
spond to  any  comparable  discontinuity  in  suitable 
habitats." 

We  have  popvilation  samples  of  Pinus  strohiformis  in 
our  arboretum  that  originated  from  Greenlee,  AZ,  at  lati- 
tude 33°39'  N.,  longitude  109°16'  W.,  elevation  2,740  m 
a.s.l.  This  stand  is  now  30  years  of  age.  Its  fertility  began 
at  the  age  of  22  years.  Its  cones  are  surprisingly  variable 
in  size,  from  the  smaller  cones  similar  to  Pinus  flexilis  to 
very  big  cones  with  apophyses,  similar  to  Pinus  aya- 
cahuite var.  brachyptera.  Such  a  variation  supports  the 
hypothesis  about  the  hybrid  origin  of  this  species. 

TESTS  OF  SPECIES  POTENTIALS 

The  response  of  a  species  to  the  impact  of  an  environ- 
mental stress  allows  us  to  recognize  intrinsic  characteris- 
tics and  appearances  that  we  do  not  usually  encounter  un- 
der normal  circumstances.  Therefore,  we  are  focused  on 
testing  the  collection  of  pine  species  at  the  arboretum 
within  the  zone  where  there  is  a  heavy  impact  of  both 
emissions  and  severe  mountain  climates. 

CONCLUSIONS 

The  threatened  trees  of  some  forest  species  imder  heavy 
emissions  often  show  a  surprising  ability  to  endure  the  se- 
lection pressure  of  the  toxic  environment.  Evidently  this 
depends  on  the  life  history  of  the  species  in  question. 
Some  conditions  under  which  we  feel  the  tested  species 
have  a  better  predisposition  to  siirvive  are: 

1.  The  species  is  of  hybrid  origin. 

2.  The  species  has  imdergone  a  long-term  interconti- 
nental migration. 

3.  The  species  has  been  subjected  to  repeated  strong 
directional  or  even  catastrophic  selection  under  the  im- 
pact of  volcanic  activity.  Volcanic  areas  are  noted  for 
their  toxic  air  and  soil  similar  in  some  ways  to  industrial 
emission  areas. 


REFERENCES 

Bannister,  M.  H.  1965.  Variation  in  the  breeding  system 
of  Pinus  radiata.  In:  Baker,  H.  G.;  Stebbins,  G.  L.,  eds. 
The  genetics  of  colonizing  species.  In:  Proc.  First  Int. 
Un.  Biol.  Sci.  Symp.  Gen.  Biol.  Davis;  1964  February. 
New  York  and  London:  Academic  Press:  353-374. 

Bigelow,  R.  S.  1965.  Hybrid  zones  and  reproductive  isola- 
tion. Evolution.  19:  449-508. 

Bouvarel,  P.  1960.  Note  sur  la  resistance  au  froid  de 
quelques  provenances  de  pin  maitime.  Revue  forestiere 
francaise.  7:  495-508. 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  Misc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture, 
Forest  Service.  97  p. 

Diels.  1910.  ex  Mirov  (1967). 

Dobzhanski,  Th.;  Ayala,  F.;  Stebbins,  L.  G.;  Valentine, 

L.  W.  1976.  Evolution.  San  Francisco:  Freeman. 
Emberger.  1944.  ex  Mirov  (1967). 

Hansen,  H.  1942.  The  influence  of  volcanic  eruptions  upon 

post-Pleistocene  forest  succession  in  central  Oregon. 

American  Journal  of  Botany.  29:  214-219. 
Hansen,  H.  1943.  Paleoecology  of  the  sand  dune  bogs  on 

the  southern  Oregon  coast.  American  Joiirnal  of  Botany. 

30:  335-340. 

Heimburger,  C.  C.  1983.  The  evolution  of  Black  spruce. 

Proc.  XDC.  Meet.  CTIA,  Part  2:  163-166.  Toronto. 
Illoff,  P.  M.,  Jr.;  Mirov,  N.  T.  1956.  Composition  of  gimi 

turpentines  of  pines.  XXV.  A  report  on  two  white  pines: 

Pinus  koraiensis  from  Korea  and  Pinus  peuce  from 

Macedonia.  Amer.  Pharm.  Ass.  J.  Sci.  Ed.  45:  77-81. 
Kahak  ,  K.  1971.  Arboretum  Sofronka.  Introduction 

results  in  1958-1968.  Praha.  38  p.  illus. 
Kahak,  K  1988.  Contribution  to  maintaining  continuity  of 

the  Norway  spruce  in  the  Ore  Moimtains.  Folia  Mus. 

Rer.  Natur.  Bohem.  Occid.  Plzai,  Botanika  27.  58  p. 
Kahak  ,  K  1991.  Colonizing  ability  of  some  introduced 

species  of  pines.  FoKa  dendrologica.  18:  187-200.  Veda, 

Bratislava. 

Lewis,  H.  1962.  Catastrophic  selection  as  a  factor  in  spe- 
ciation.  Evolution.  16:  257-271. 

Lewontin,  R.  C;  Birch,  L.  C.  1966.  Hybridization  as  a 
source  of  variation  for  adaptation  to  new  environments. 
Evolution.  20:  315-336. 

Mayr,  E.  1942.  Systematics  and  the  origin  of  species.  New 
York:  Columbia  University  Press. 

Mayr,  E.  1965.  Summary.  In:  Baker,  H.  G.;  Stebbins,  G. 
L.,  eds.  The  genetics  of  colonizing  species.  In:  Proc.  First 
Int.  Un.  Biol.  Sci.  Symp.  Gen.  Biol.  Davis;  1964  Febru- 
ary. New  York  and  London:  Academic  Press:  553-562. 

Mayr,  E.  1979.  Evolution  xmd  Vielfalt  des  Lebens. 
Springer-Verlag. 

Mirov,  N.  T.  1967.  The  Genus  Pinus.  New  York:  Ronald  Press. 

Morgenstem,  E.  K.;  Farrar,  J.  L.  1964.  Introgressive 
hybridization  in  red  spruce  and  black  spruce.  Tech. 
Rep.  4.  Faculty  of  Forestry,  Univ.  of  Toronto. 

Mottl,  J.,  Prudic,  Z.  1982.  Analyza  riistu  limby  na 
Klmovci — V  Krusnych  horach  a  jejf  roubavanci. 
Tschechisch.  Zpravy  les.  vyzkumu  27(4):  9-13. 

Parsons,  P.  A.  1963.  Migration  as  a  factor  in  natural  selec- 
tion. Genetica.  33:  184-206. 

Raven,  P.  1964.  Catastrophic  selection  and  edaphic 
endism.  Evolution.  17:  336-338. 


129 


COMPETITION  AND  CROWN 
CHARACTERISTICS  OF  WHITEBARK 
PINE  FOLLOWING  LOGGING  IN 
MONTANA,  U.S.A. 

Todd  Kipfer 
Katherine  Hansen 
Ward  McCaughey 


Abstract — This  study  was  designed  to  evaluate  crown  develop- 
ment of  whitebark  pine  (Pinus  albicaulis)  subjected  to  intertree 
competition  in  young  mountainous  stands  in  the  Yellowstone 
area.  A  distance-dependent  competition  index  was  used  to  assess 
intertree  competitive  effects  on  crown  growth.  Significant  in- 
verse correlations  were  found  between  the  competition  index  and 
height,  crown  diameter,  crown  diameter/height,  and  crown  vol- 
ume; however,  only  20  percent  of  the  variation  was  explained  by 
measured  variables.  Competition  thresholds,  indicating  when 
crown  growth  becomes  limited,  were  estimated.  In  future  stud- 
ies, crown  vigor  or  density  measures  could  perhaps  yield  a  better 
competition  indicator. 


This  study  addresses  the  response  of  regenerating 
whitebark  pine  {Pinus  albicaulis)  to  competition  following 
logging.  Human  activities,  such  as  logging,  in  the  movm- 
tainous  environments  of  western  North  America  are  cre- 
ating a  variety  of  impacts  at  both  the  individual  scale  and 
at  the  ecosystem  scale.  Concern  about  regeneration 
crown  growth  and  future  cone  production  in  logged  sites 
prompted  this  study.  Cone  production  is  a  product  of 
mean  shoot  production  multiplied  by  the  number  of  fertile 
shoots  (Weaver  and  Forcella  1986).  Effects  of  competition 
were  hypothesized  to  affect  crown  development  of  white- 
bark pine.  Intertree  competition  was  measured  on  indi- 
vidual, regenerating  whitebark  pine  in  logged  sites  in 
southwestern  Montana  near  Yellowstone  National  Park. 
This  study  is  part  of  a  larger  research  project  conducted 
by  Kipfer  (1992). 

Whitebark  pine  are  found  on  sites  with  a  wide  range 
of  geologic,  geomorphic,  and  ecologic  conditions  (Hansen- 
Bristow  and  others  1990).  Their  growth  characteristics 
depend  on  numerous  factors,  including  competition,  that 
vary  spatially  and  temporally.  Quantifying  the  influence 
of  competing  vegetation  on  conifer  grov^i^h  may  help  to  de- 
termine silvicultural  prescriptions  such  as  thinning  inten- 
sity, optimize  management  decisions,  and  provide  a  basis 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Todd  Kipfer  is  a  Doctoral  Student,  Department  of  Geography,  Arizona 
State  University,  Tempe,  AZ  85281;  Katherine  Hansen  is  an  Associate 
Professor  of  Geography,  Department  of  Earth  Sciences,  Montana  State 
University,  Bozeman,  MT  59717;  Ward  McCaughey  is  a  Research  For- 
ester, USDA  Forest  Service,  Intermountain  Research  Station,  Forestry 
Sciences  Laboratory,  Bozeman,  MT  59717. 


for  growth  models  of  young  forest  stands  (Wagner  and 
Radosevich  1991b).  For  example,  one  management  option 
for  whitebark  pine  is  to  increase  or  sustain  seed  produc- 
tion for  regeneration  and  for  a  food  source  for  the  Clark's 
nutcracker  (Nucifraga  columbiana),  the  red  squirrel 
{Tamiasciurus  hudsonicus),  the  grizzly  bear  (Ursus 
arctos  horribilis),  and  a  host  of  other  birds  and  mammals 
(Craighead  and  others  1982;  Kendall  1983;  Kendall  and 
Amo  1990;  McCaughey  and  Schmidt  1990).  Seed  produc- 
tion may  increase  with  v^dder,  more  diflFuse  crowns  (Eggers 
1986;  Spurr  and  Barnes  1980),  and,  therefore,  silvicultural 
prescriptions  that  increase  crown  diffiisivity  are  desirable. 
The  effects  of  silvicultural  management  options  for  main- 
taining or  enhancing  whitebark  pine  ecosystems  have  had 
limited  study  (Eggers  1990;  Schmidt  and  McCaughey 
1990). 

PREVIOUS  STUDIES 

Whitebark  pine  is  considered  relatively  intolerant  of  com- 
petition and  shade  (Amo  and  Weaver  1990),  although  the 
trees  may  be  more  shade  tolerant  at  an  early  stage  and 
less  tolerant  at  later  developmental  stages  (McCaughey 
and  Schmidt  1990).  Although  initial  observations  by 
Eggers  (1990)  indicate  that  suppressed  whitebark  pine 
seedlings  and  saplings  respond  little  to  the  removal  of  for- 
est competition,  the  effect  of  competition  on  postlogging 
regeneration  needs  further  study.  Weaver  and  others 
(1990)  assessed  competition  in  terms  of  stand  develop- 
ment and  hypothesized  that  although  the  density  of 
whitebark  pine  seedlings  remained  approximately  con- 
stant across  stands  of  different  ages,  growth  from  seedling 
to  sapling  size  could  not  be  supported  in  stands  older  than 
100  years.  They  suggested  that  clearings  needed  to  be 
larger  than  10  m^  for  seedlings  to  grow  to  sapling  size. 

The  effect  of  competition  on  the  growth  of  individual 
trees  has  been  studied  for  many  economically  important 
trees  in  plantations  or  homogeneous  stands  (Daniels  and 
others  1986;  Tome  and  Burkhart  1989;  Wagner  and 
Radosevich  1991a,b).  These  studies  have  primarily  used 
individual  tree  grovvi:h  models  based  on  competition  indi- 
ces and  have  shown  an  inverse  relationship  between  the 
amount  of  competition  from  neighbors  and  individual  tree 
growth  performance  (Daniels  1976;  Daniels  and  others 
1986;  Lorimer  1983;  Tome  and  Burkhart  1989).  This 
study  focused  on  crown  development  of  whitebark  pine 
as  related  to  a  distance-dependent  competition  index. 
Managers  could  use  this  index  to  determine  the  optimum 


130 


spacing  or  number  and  spatial  arrangement  of  competi- 
tors for  maximimi  crown  development  of  whitebark  pine, 

METHODS 

Three  logged  stands  were  sampled  in  the  Gallatin  Na- 
tional Forest  in  southwestern  Montana  near  Yellowstone 
National  Park.  The  stands  ranged  in  elevation  from  an 
average  of  2,260  m  at  Moose  Creek  to  2,475  m  and  2,535  m 
at  Teepee  Creek  1  and  2,  respectively.  The  three  stands 
had  been  clearcut  between  1968  and  1972.  Data  collec- 
tion focused  upon  intra-  and  intertree  characteristics  for 
whitebark  pine  and  neighboring  trees.  One  hundred 
whitebark  pine  individuals  (more  than  0.5  m  tall)  were 
identified  within  each  of  the  three  stands  by  a  systematic- 
random  sampling  procedure  along  transects.  Three 
different-sized  circular  plots  were  nested  around  each  in- 
dividual whitebark  pine.  Within  each  circular  plot,  trees 
of  specific  sizes  were  identified  as  competitors  using  the 
following  criteria:  within  the  3-m-radius  plot,  all  tree 
seedlings  and  saplings  were  identified  as  competitors; 
within  the  6-m-radius  plot  all  trees  with  a  diameter  at 
breast  height  (d.b.h.)  greater  than  or  equal  to  4  cm  and 
less  than  10  cm  were  identified  as  competitors;  and  within 
the  9-m-radius  plot  all  trees  with  d.b.h.  greater  than  or 
equal  to  10  cm  were  identified  as  competitors.  This 
method  approximated  a  fixed  angle  gauge  sweep  (such  as 
that  used  by  Tome  and  Burkhart  1989)  and  requires  that 
trees  farther  away  from  the  individual  tree  be  larger  to  be 
considered  competitors. 

Five  attributes  were  measured  for  each  whitebark  pine: 
(1)  age,  (2)  height,  (3)  diameter  at  breast  height  (d.b.h.), 
(4)  average  live  crown  diameter,  and  (5)  height  of  the 
maximum  crown  diameter  above  the  ground  surface. 
Crown  and  height  variables  were  measured  to  the  nearest 
0.1  m  and  d.b.h.  to  the  nearest  0.1  cm. 

Total  sample  size  was  reduced  from  300  to  220  after 
tree  ages  showed  that  some  selected  whitebark  pine  indi- 
cated some  had  established  prior  to  logging  (advance  re- 
generation). It  is  difficult  to  evaluate  intertree  competi- 
tion prior  to  logging,  and  these  were  therefore  eliminated 
from  further  study.  A  ratio  of  crown  diameter  to  tree 
height  iCD/H)  was  computed  to  reflect  the  degree  of  hori- 
zontal crown  growth.  Crown  volume  was  computed  using 
the  cone  volume  equation:  (Volume  =  1/3*  PI  *  R'^  *  H). 
Measurements  on  competing  trees  within  the  9-m-radius 
plot  were  tree  species,  total  tree  height,  diameter  at 
breast  height,  and  distance  (measured  to  nearest  0.1  m) 
to  the  subject  whitebark  pine. 

A  distance-weighted  size-ratio  index  or  distance- 
diameter  index  (Alemdag  1978;  Daniels  1976;  Hamilton 
1969;  Hegyi  1974;  Tome  and  Burkhart  1989)  was  selected 
to  develop  a  competition  index,  from  which  intertree  com- 
petition could  be  quantified.  The  distance-diameter  index 
is  a  distance-dependent  index  that  svuns  a  size  ratio  of 
diameters  (d.b.h.  or  basal  diameter)  of  competing  trees  to 
the  diameter  of  a  subject  tree.  The  general  theory  is  that 
larger  competing  trees,  relative  to  the  individual  tree, 
contribute  more  competition  than  do  smaller  trees.  The 
index  also  assumes  that  competitive  influence  decreases 
with  increasing  distance  between  the  individual  tree  and 
the  competitor. 


The  distance-diameter  index  was  modified  for  white- 
bark pine  because  the  trees  often  had  multiple  stems  or 
did  not  reach  breast  height.  Height  was  substituted  for 
diameter  and  distance  weighting  functions  were  applied 
to  develop  the  modified  competition  index: 

N 

C.I.=  Z  (if./i/.)*function(Z)/ST..) 
i=l    '  ' 

where  C.I.  is  the  competition  index;  if.  is  the  tree  height 
of  the  subject  tree;  H.  is  the  tree  height  of  the  competing 
tree;  and  DIST.^  is  a  distance  decay  function.  Three  dis- 
tance decay  functions  were  used:  {1/DIST  ,  l/DIST^  ,  and 
e-°'s^y). 

Regression  analysis  was  used  to  evaluate  the  relation- 
ship between  competition  and  crown  characteristics  of  the 
whitebark  pine.  The  three  distance-decay  functions  were 
used  as  dependent  variables  and  tree  height  and  three 
crown  measures  (crown  diameter,  crown  diameter  to  tree 
height  ratio,  and  crown  volimie)  as  independent  variables. 

RESULTS 

Competition  index  values  varied  according  to  the  form 
of  the  distance-decay  function  used  for  whitebark  pine. 
Index  values  calculated  using  1/distance  and  1/distance^ 
were  similar  with  mean  values  ranging  from  13.12  to 
17.12.  Index  values  using  the  e~^'^^'^^  decay  function 
ranged  from  3.11  to  4.22. 

Tree  heights  of  whitebark  pine  were  significantly  corre- 
lated (P  =  0.007)  with  the  competition  index;  however, 
values  for  the  three  distance-decay  functions  (0.033  for 
1/distance,  0.027  for  l/distance^,  and  0.032  for  e"^''*^^^) 
indicate  that  the  linearity  of  that  relationship  is  poor  as 
shown  for  1/distance  in  figure  1.  The  intercept  value  in 
the  regression  equation  (y  =  a  +  bx)  was  2.047,  which  may 
represent  maximimi  tree  heights  for  whitebark  pine  in 
the  age  classes  represented  by  this  study,  when  growing 
in  the  absence  of  competition. 

Crown  diameters  of  whitebark  pine  were  significantly 
correlated  {P  =  0.000)  with  the  competition  index,  shown 
for  1/distance  in  figure  1,  with  R^  values  varying  depend- 
ing on  the  form  of  the  distance-decay  function  (0.131  for 
1/distance,  0.096  for  l/distance^,  and  0.128  for  e"^*^^^^). 
The  intercept  value  (0.820)  of  the  regression  equation 
poorly  represents  the  maximum  possible  crown  diameter, 
since  crown  diameter  values  above  1.0  m  were  common. 

The  crown  diameter  to  tree  height  ratio  and  crown  vol- 
ume were  significantly  correlated  to  competition  index  as 
shown  for  the  1/distance  decay  function  in  figure  1.  Re- 
gression R^  values  for  crown  diameter  to  tree  height  ratio 
varied  depending  on  the  form  of  the  distance-decay  func- 
tion in  the  competition  index  equation  (0.122  for  1/distance, 
0.079  for  l/distance^,  and  0.120  for  e-^'^'=^).  Similar  regres- 
sion i?^  values  were  obtained  in  correlating  competition 
index  with  crown  volume  using  the  three  forms  of  the 
distance-decay  function  (0.092  for  1/distance,  0.058  for 
l/distance^  and  0.091  for  e-^^^-^^). 

The  1/distance  decay  function  in  the  competition  index 
model  yielded  slightly  higher  R  ^  values  for  tree  height, 
crown  diameter,  crown  diameter  to  tree  height  ratio,  and 
crown  voliraie  of  whitebark  pine.  The  1/distance  decay 


131 


S.0- 
4.5' 
4.0 

3.5  ■ 

ao 
^6 

2.0 

1J 

1.0 

as 
ao 


Tro  halght  »  2.047  -  0.0133'C.I. 
R-squarod  >  0.033 
p  =  0.007  n  >  220 


•  •       •  • 

£  li  *•*  t  .y  .  r 


* 


—  IX)- 


0.0 


Crown  diameter  >  0.820  -  0.0117*C.I. 
R-tquarad  =  0.131 
p  -  0.000  n  -  220 


•  a     •  a 

•  aa  « 

a  a 

•aaa  a  a  a  a 


aaa   •  a  •      a  a 
■       a  aa  a  a 
a        a  a 

a         a  a  aa 
•a  a    «a  a 


a  aa  •aaa 

•aa  a  aa  a« 

a     a        a  a 

aaa  a  aa  • 

a        aa>*  a 

a  a  aa  • a 

a  aaa 


a  a 
a  a 


a  a  a 
a  a 


Competition   Index  (I/distance) 


Competition  index  (l/dietance) 


a 

a 

Crown  diameter/Tree  height  «  0.428  -  0.00481 'CI. 
R-squarad  '  0.122 
p  -  0.000  n  3  220 

a 

Sa          •    •  a 
a'«       a-a  a 
J  a        a       •  •• 

a 

aaa       a  • 
a^    ^'V.-.a  ;C 

a  • 
•      a  •  •                      •  a 

a     a  a 

•  ^"'^ 

a 

aa 

• 

10  20  30  40  60 

Competition   index  (I/distance) 


60 


CO 

< 


Crown  vokima  <s  0.300  -  0.00056*0.1. 
R-aquarad  •  0.002 
p  «  0.000  n  =  220 


#  • 


•  t  • 


•••'a  *. 


10  20  30  40  SO 

Competition   index  (1/distance) 


Figure  1 — Bivariate  scatterplots  of  total  tree  heiglit,  crown  diameter,  crown  diameter/lielght,  and  crown 
volume  of  whiitebarl^  pine  plotted  against  competition  index  using  a  distance-decay  function  of  1/distance. 
Regression  values  shown  are  model  and  model  coefficients,  fl-square,  p-level,  and  sample  size  (n).  (Data 
from  all  tiiree  study  sites — iy/loose  Creek,  Teepee  Creek  1 ,  and  Teepee  Creek  2.) 


function  was  used  for  development  of  a  potential  competi- 
tion threshold  because  it  was  consistently  a  stronger  pre- 
dictor of  the  four  growth  variables. 

Scatterplots  were  evaluated  from  each  study  site  for 
competition  index  values  above  which  there  was  visually 
less  variability  in  the  data.  No  identifiable  threshold  was 
observed  for  whitebark  pine  in  the  Moose  Creek  stand, 
and  therefore,  in  order  to  increase  the  sample  size,  the 
two  Teepee  Creek  stands  were  combined.  Potential  com- 
petition thresholds  were  estimated  for  total  tree  height, 
crown  diameter,  crown  diameter  to  tree  height  ratio,  and 
crown  voltmie  for  sampled  whitebark  pine  in  the  two  Tee- 
pee Creek  stands  (fig.  2).  Variability  of  data  below  poten- 
tial competition  thresholds  indicates  that  tree  growth  fac- 
tors may  not  be  limiting.  The  majority  of  whitebark  had 
competition  index  values  below  the  estimated  threshold 
values. 

Regression  analysis  of  the  competition  index  against 
the  four  whitebark  pine  measures  using  only  those  trees 
greater  than  the  threshold  values  jdelded  poor  results  for 


tree  height  (R^  =  0.128,p  =  0.062,  n  =  27),  crown  diameter 
(R^  =  0.146,  p  =  0.106,  n  =  18),  and  crown  volume  (R^  = 
0.091,  p  =  0.120,  n  =  27).  A  statistically  significant  corre- 
lation, however,  was  found  between  the  competition  index 
and  the  crown  diameter  to  tree  height  ratio  (CD/H  = 
0.846  -  0.0142*C.I.:  R^  =  0.230, p  =  0.000,  n  =  18). 

DISCUSSION 

The  specific  competition  index  values  developed  by  this 
study  explain  approximately  20  percent  of  the  variation  in 
competitive  pressure  exerted  on  whitebark  pine  from  the 
numbers,  sizes,  and  spatial  patterns  of  neighboring  trees. 
Assessment  of  competition  on  crown  development  of 
whitebark  pine  is  difficult  when  data  come  from  present 
spatial  relationships  and  past  growing  conditions  are  un- 
known. Competition  was  a  significant  factor  influencing 
crown  characteristics  of  whitebark  pine  in  this  study.  It 
is  important  to  note  that  these  results  only  show  that 
competition  is  a  significant  factor  influencing  whitebark 


132 


5.0 
4.5 
4.0 

3.5 

ao 

2.5 
2.0 
1.5 
1.0 
OJ 

ao 


• 

• 

• 

• 

<               Potential  threshold 

• 

;i.r  J»  •  t,  - 

•  •  • 

•  •  • 

••  • 

*  *  • 

• 

•••      •      ••  • 

•    •  • 

•  • 

20  30  40  SO 

Competition    index  (1/distance) 


E 


OX 


(  •  •••• 


•  •  •  •  •  I 


Potential  threshold 


10  20  30  40  50 

Competition   Index  (1/dlstance) 


Potential  threshold 

• 

• 

<  

• 

\,     •  •  ' 

•  •  •  •       •  • 

•  • 

m  • 
• 

• 

• 

• 

•    •  • 
m 

•  • 

20  30  40  60 

Compstltlon  Index  (1/dlstance) 


n 
< 

o 

E 

o 
> 


•  •  •/ 


Potential  threshold 


10  20  30  40  £ 

Competition   index  (I/distance) 


Figure  2 — Potential  competition  thresholds  for  bivariate  scatterplots  of  total  tree  height,  crown  diameter, 
crown  diameter/height,  and  crown  volume  of  whitebark  pine  plotted  against  competition  index  using  a 
distance-decay  function  of  1/distance.  (Data  from  Teepee  Creek  1  and  Teepee  Creek  2  only.) 


pine.  The  mechanisms  of  this  competitive  influence  are 
unknown. 

The  distance-decay  function  1/distance  in  the  competi- 
tion index  model  created  an  expected  inverse  relationship 
of  tree  growth  measure  to  index.  As  distance  from  the 
subject  tree  increases  effects  from  competition  trees  de- 
crease. Tree  heights  and  crown  volumes  had  the  poorest 
correlations  with  the  competition  index.  Poor  correlations 
with  crown  volume  may  be  due  to  sampHng  procedures. 
In  future  studies,  crown  vigor  and  crown  biomass  meas- 
urements (stem  spacing)  might  be  used  to  weight  crown 
volume  measures,  perhaps  yielding  a  better  indication  of 
the  effect  of  competition  on  growth. 

Crown  diameters  and  crown  diameter  to  tree  height  ra- 
tios had  good  correlations  with  the  competition  index. 
Still,  only  12  to  13  percent  of  the  variation  in  these  meas- 
ures was  explained  by  the  competition  index.  The  addi- 
tion of  age  and  an  indicator  variable  for  stand  improved 
this  relationship  for  the  crown  diameter  measure,  increas- 
ing the  R  2  value  to  0.250.  It  did  not,  however,  yield  a  sig- 
nificant correlation  for  the  crown  diameter  to  tree  height 


ratio.  The  crown  diameter  to  tree  height  ratio  may  repre- 
sent a  characteristic  crown  shape  for  young  whitebark 
pine. 

Poor  correlations  between  whitebark  pine  crown  charac- 
teristics and  competition  index  may  be  due  to  factors  in- 
fluencing individual  tree  size  (local  density,  plant  geno- 
type, seed  size,  emergence  time,  microhabitat  variations, 
and  unknown  historical  growing  conditions).  A  competi- 
tion index  coxdd  incorporate  a  variety  of  growing  condi- 
tions over  time  and  the  use  of  incremental  growth  charac- 
teristics of  annual  growth  could  provide  some  improvement 
without  substantial  measurement  efforts. 

Potential  competition  thresholds  estimated  for  the  two 
Teepee  Creek  sites  may  be  artifacts  due  to  relatively  few 
trees  with  high  competition  values,  or  they  may  be  actual 
values  of  competitive  pressure  indicating  when  growth 
becomes  limited  by  competition.  Multiple-aged  stands 
with  varying  tree  densities  should  be  sampled  in  future 
studies.  To  compare  stands  of  different  ages,  the  age- 
independent  competition  index  methods  of  Lorimer  (1983) 
may  provide  a  strong  starting  point.  Although  regression 
analysis  can  comparatively  evaluate  the  influences  of 


133 


competitive  pressure  on  whitebark  pine  characteristics, 
actual  relationships  are  more  complex  than  the  linear  re- 
gression models  might  suggest.  Nonlinear  regression 
methods  should  be  explored  for  evaluating  competition 
influences. 

Timber  harvests  provide  an  important  but  declining 
part  of  our  regional  economy  (Powers  1991),  and  economic 
considerations  are  an  important  component  of  forest  man- 
agement policies.  Timber  harvest  can  cause  accelerated 
slope  failures,  erosion,  and  stream  sedimentation  from 
roads  (Marston  and  Anderson  1991)  when  improperly 
done.  Road  building  in  association  with  logging  provides 
access  to  whitebark  pine  regeneration  sites;  however,  it  is 
not  believed  that  logging  will  be  a  major  means  for  regen- 
erating whitebark  pine  (Arno  1986). 

It  is  imperative  to  find  viable  management  options  for 
increasing  crown  growth  and  the  potential  for  increased 
cone  production  for  whitebark  pine  because  of  its  impor- 
tance to  the  endangered  grizzly  and  many  other  wildlife 
species  in  the  Yellowstone  ecosystem.  The  importance  of 
quantifying  how  competing  trees  influence  crown  charac- 
teristics of  whitebark  pine  in  logged  stands  comes  from  a 
need  to  increase  our  knowledge  of  how  competition  affects 
crown  development  of  whitebark  pine. 

ACKNOWLEDGMENT 

This  research  was  supported  in  part  by  funds  provided 
by  the  Intermoxmtain  Research  Station,  Forest  Service, 
U.S.  Department  of  Agricultiu*e. 

REFERENCES 

Alemdag,  I.  S.  1978.  Evaluation  of  some  competition  indi- 
ces for  the  prediction  of  diameter  increment  in  planted 
white  spruce.  Inf.  Rep.  FMR-X-108.  Canadian  Forest 
Service  Forest  Management  Institute.  39  p. 

Arno,  S.  F.  1986.  Whitebark  pine  cone  crops — a  diminish- 
ing source  of  wildlife  food?  Western  Journal  of  Applied 
Forestry.  1(3):  92-97. 

Arno,  S.  F.;  Weaver,  T.  1990.  Whitebark  pine  community 
types  and  their  patterns  on  the  landscape.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposium  on  whitebark  pine  ecosystems:  ecology 
and  management  of  a  high-mountain  resource;  1989 
October  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  97-105. 

Craighead,  J.  J.;  Svunner,  J.  S.;  Scaggs,  G.  B.  1982.  A  de- 
finitive system  for  analysis  of  grizzly  bear  habitat  and 
other  wilderness  resources.  Wildlands  Inst.  Monogr.  1. 
Missoula,  MT:  University  of  Montana.  279  p. 

Daniels,  R.  F.  1976.  Simple  competition  indices  and  their 
correlation  with  annual  loblolly  pine  tree  growth.  Forest 
Science.  22:  454-456. 

Daniels,  R.  F.;  Burkhart,  H.  E.;  Clason,  T.  R.  1986.  A 
comparison  of  competition  measures  for  predicting 
growth  of  loblolly  pine  trees.  Canadian  Journal  of  For- 
est Research.  16:  1230-1237. 


Eggers,  D.  E.  1986.  Management  of  whitebark  pine  as  po- 
tential grizzly  bear  habitat.  In:  Contreras,  Glen  P.; 
Evans,  Keith  E.,  comps.  Proceedings — grizzly  bear  habi- 
tat symposiimi;  1985  April  30-May  2;  Missoula,  MT. 
Gen.  Tech.  Rep.  INT-207.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  170-175. 

Eggers,  D.  E.  1990.  Silvicultiiral  management  alterna- 
tives for  whitebark  pine.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  maneige- 
ment  of  a  high-mountain  resource;  1989  October  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  324-328. 

Hamilton,  G.  J.  1969.  The  dependence  of  volume  incre- 
ment of  individual  trees  on  dominance,  crown  dimen- 
sions, and  competition.  Forestry.  42:  133-144. 

Hansen-Bristow,  K.;  Montague,  C;  Schmid,  G.  1990. 
Geology,  geomorphology,  and  soils  within  whitebark 
pine  ecosystems.  In:  Schmidt,  W3anan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  October  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  62-71. 

Hegyi,  F.  1974.  A  simulation  model  for  managing  jack- 
pine  stands.  In:  Fries,  J.,  ed.  Growth  models  for  tree 
and  stand  simulation.  Res.  Notes  30.  Stockholm:  Royal 
College  of  Forestry:  74-90. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  black  and  grizzly 
bears  in  the  Yellowstone  area.  In:  Bears — their  biology 
and  management;  International  Conference  on  Bear 
Research  and  Management.  Madison,  WI:  International 
Bears  Association.  5:  166-173. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an 
important  but  endangered  wildlife  resource.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposiima  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain 
resource;  1989  October  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
264-273. 

Kipfer,  T.  R.  1992.  Post-logging  stand  characteristics  and 
crown  development  of  whitebark  pine  (Pinus  albicau- 
lis).  Bozeman,  MT:  Montana  State  University,  Depart- 
ment of  Earth  Science.  83  p.  Thesis. 

Lorimer,  C.  G.  1983.  Tests  of  age-independent  competition 
indices  for  individual  trees  in  natural  hardwood  stands. 
Forest  Ecology  and  Management.  6:  343-360. 

Marston,  R.  A.;  Anderson,  J.  E.  1991.  Watersheds  and 
vegetation  of  the  Greater  Yellowstone  Ecosystem.  Con- 
servation Biology.  5(3):  338-346. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposivun  on 
whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource;  1989  October  29-31; 


134 


Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  85-96. 

Powers,  T.  M.  1991.  Ecosystem  preservation  and  the 
economy  in  the  Greater  Yellowstone  Area.  Conservation 
Biology.  5(3):  395-404. 

Schmidt,  W.  C.;  McCaughey,  W.  W.  1990.  Whitebark 
pine — a  subalpine  species  needing  silvicultural  atten- 
tion. In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  October  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
tiu-e,  Forest  Service,  Intermountain  Research  Station: 
373-374. 

Spurr,  S.  H.;  Barnes,  B.  V.  1980.  Forest  ecology.  3d  ed. 
New  York:  John  Wiley  and  Sons.  687  p. 

Tome,  M.;  Burkhart,  H.  E.  1989.  Distance-dependent  com- 
petition measures  for  predicting  growth  of  individual 
trees.  Forest  Science.  35(3):  816-831. 

Wagner,  R.  G.;  Radosevich,  S.  R.  1991a.  Neighborhood 
predictors  of  interspecific  competition  in  young 


Douglas-fir  plantations.  Canadian  Journal  of  Forest  Re- 
search. 21:  821-828. 
Wagner,  R.  G.;  Radosevich,  S.  R.  1991b.  Interspecific  com- 
petition indices  and  other  factors  influencing  the  perfor- 
mance of  Douglas-fir  saplings  in  the  Oregon  Coast 
Range.  Canadian  Journal  of  Forest  Research.  21:  829- 
835. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1986.  Cone  production 
in  Pinus  albicaulis  forests.  In:  Shearer,  R.  C,  comp. 
Proceedings — conifer  tree  seed  in  the  inland  mountain 
west  symposium;  1985  August  5-6;  Missoula,  MT.  Gen. 
Tech.  Rep.  INT-203.  Ogden,  UT:  U.S.  Department  of 
Agriciilture,  Forest  Service,  Intermotmtain  Research 
Station:  68-76. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1990.  Stand  develop- 
ment in  whitebark  pine  woodlands.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposimn  on  whitebark  pine  ecosystems:  ecology 
and  management  of  a  high-mountain  resource;  1989 
October  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agricvilture,  Forest  Ser- 
vice, Intermountain  Research  Station:  151-155. 


135 


FIRE  ECOLOGY  OF  WHITEBARK  PINE 
FORESTS  OF  THE  NORTHERN  ROCKY 
MOUNTAINS,  U.SA. 

Penelope  Morgan 
Stephen  C.  Bunting 
Robert  E.  Keane 
Stephen  F.  Amo 


Abstract— Fires  once  occurred  at  intervals  between  30  and  300 
years  in  whitebark  pine  (Pinus  albicaulis)  forests  in  the  Northern 
Rocky  Mountains,  U.S.A.,  but  since  the  early  1900's  fewer  fires 
have  occurred,  contributing  to  declining  abundance  of  whitebark 
pine.  In  the  absence  of  fire  or  other  major  disturbance,  whitebark 
pine  is  replaced  by  other  conifers  on  most  of  the  upper  subalpine 
landscape.  Whitebark  pines  often  survive  low-intensity  surface 
fires.  Large  stand-replacement  fires  also  benefit  this  species  by 
creating  the  open,  burned  sites  where  regeneration  is  most 
successful. 


In  the  Northern  Rocky  Mountains  of  western  North 
America,  whitebark  pine  (Pinus  albicaulis)  historically 
dominated  many  upper  subalpine  forests.  These  high- 
elevation  forests  usually  have  poorly  developed,  rocky 
soils  and  are  often  located  within  wilderness  or  roadless 
areas.  As  a  consequence,  whitebark  pine  is  seldom  har- 
vested for  forest  products,  but  it  is  important  for  scenic, 
watershed,  and  wildlife  habitat  values. 

Whitebark  pine  dominates  middle-  and  late-successional 
stages.  In  the  absence  of  major  disturbance,  however, 
whitebark  pine  is  eventually  replaced  by  the  more  shade- 
tolerant  subalpine  fir  (Abies  lasiocarpa)  and  Engelmann 
spruce  (Picea  engelmannii)  in  most  of  its  range  in  the  upper 
subalpine  forest  zone  of  the  Northern  Rocky  Mountains 
(fig.  1).  Whitebark  pine  is  a  common  serai  component 
of  upper  subalpine  forests  found  on  the  Abies  lasiocarpa/ 
Vaccinium  scoparium,  A.  lasiocarpa/Luzula  hitchcockii, 
and  A.  lasiocarpa/Arnica  cordifolia  habitat  types  (Pfister 
and  others  1977;  Steele  and  others  1981,  1983;  Weaver 
and  Dale  1974).  These  sites  are  cold,  with  July  mean 
temperatures  averaging  13  to  15  °C  (Pfister  and  others 
1977).  Although  annual  precipitation  averages  610  to 
1,780  mm,  summer  drought  is  common  (Arno  and  Hoff 
1989;  Pfister  and  others  1977;  Weaver  and  Dale  1974). 


Whitebark  pine  is  also  found  in  pure  stands  on  relatively 
dry  and  severe,  windswept  sites  near  timberline  where  it 
is  the  climax  tree  species  (Amo  and  HofF  1989).  It  is  the 
sole  climax  tree  species  on  Pinus  albicaulis  habitat  types 
in  Montana,  central  Idaho,  and  western  Wyoming,  and  in 
southern  Canada  (Amo  and  Hoff  1989;  Steele  and  others 
1981, 1983).  Whitebark  pine  and  subalpine  fir  are  climax 
codominants  on  the  Pinus  albicaulis-Abies  lasiocarpa  habi- 
tat types  where  subalpine  fir  growth  is  stunted  in  the  severe 
microclimate  (Pfister  and  others  1977).  Climax  whitebark 
pine  forests  are  usually  open,  with  smcdl  patches  of  trees 
of  mixed  ages  interspersed  with  meadows  (fig.  2).  Average 
July  mean  temperatures  are  10  to  12  °C  with  severe  sum- 
mer droughts  and  frosts;  the  annual  precipitation  of  71  to 
153  cm  falls  mostly  as  snow  (Arno  and  Hoff  1989;  Pfister 
and  others  1977). 

Fires  are  very  important  to  regeneration  and  survival  of 
whitebark  pine  on  sites  where  it  is  serai.  Whitebark  pine 
often  survives  these  low-intensity  surface  fires,  which  more 
easily  kill  associated  conifers  (fig.  3).  Stand-replacing  fires 
also  benefit  whitebark  pine,  for  although  all  trees  are  usu- 
ally killed,  whitebark  pine  regenerates  on  burned  sites  more 
successfully  than  many  associated  tree  species  (Tomback 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Penelope  Morgan  is  Associate  Professor,  Department  of  Forest  Re- 
sources, University  of  Idaho,  Moscow,  ID  83843;  Stephen  C.  Bunting  is 
Professor,  Department  of  Range  Resources,  University  of  Idaho,  Moscow, 
ID  83843;  Robert  E.  Keane  and  Stephen  F.  Amo  are  Research  Ecologist 
and  Research  Forester,  Intermountain  Fire  Sciences  Laboratory,  Inter- 
mountain  Research  Station,  Forest  Service,  U.S.  Department  of  Agricul- 
ture, P.O.  Box  8089,  Missoula,  MT  59801. 


Figure  1 — Subalpine  fir  and  Engelmann  spruce 
now  dominate  many  stands  and  whole  landscapes 
where  whitebark  pine  was  historically  abundant. 
Whitebark  pine  is  declining  in  abundance  even 
where  blister  rust  is  uncommon,  such  as  on  this 
site  east  of  Yellowstone  National  Park  in  Wyoming. 


136 


and  others  1990)  (fig.  4).  Large  stand-replacing  fires  are 
infi-equent,  usually  occurring  only  during  windy  conditions 
after  prolonged  drought.  The  fires  that  burned  in  and 
around  Yellowstone  National  Park  in  1988  were  spectacu- 
lar examples  of  the  large,  high-intensity  fires  that  periodi- 
cally burn  within  whitebark  pine  and  adjacent  forest  types. 
These  fires  burned  whitebark  pine  habitats  in  a  patchy, 
stand-replacing  manner. 

During  the  10,000  years  since  development  of  forests  af- 
ter the  last  glacial  retreat,  fires  have  had  a  major  influence 
on  the  structure  and  composition  of  forests  in  the  Northern 
Rocky  Mountains  (Arno  1980).  Fire  occurrence  has  been 
significantly  altered  by  human  activity.  Native  Americans 
used  fire  to  manipulate  vegetation,  for  hunting,  for  commu- 
nication, and  for  other  purposes  (Pyne  1982).  More  recently, 
humans  have  affected  fire  occiurence  by  purposefully  or  in- 
advertently igniting  fires,  through  fire  suppression,  and  by 
grazing  domestic  livestock,  logging,  or  otherwise  altering 
the  fuels  available  to  burn.  Efi"orts  to  suppress  fires  have 
become  increasingly  effective  since  about  1935  (Arno  1980; 
Pyne  1982). 


Figure  3 — Whitebark  pine  trees  often  survive  fires  and  individual  trees  may  survive  multiple 
fires.  Information  on  fire  frequency  and  effects  is  derived  from  dates  of  fire  scars  on  tree 
sections  such  as  this  one  from  a  208-year-old  whitebark  pine  tree  in  the  Shoshone  National 
Forest  in  northwestern  Wyoming.  First  scarred  by  fires  when  only  12.2  cm  in  diameter  at 
the  base,  it  was  again  scarred  60  years  later.  It  was  cut  in  1 988, 1 1 0  years  later. 


Figure  2 — Whitebark  pine  often  is  the  only  tree 
that  can  grow  on  harsh  sites.  Here,  in  the  Challis 
National  Forest  in  central  Idaho,  whitebark  pine  is 
the  climax  tree  species. 


137 


Figure  4 — Young  whitebark  pine  trees  regener- 
ated successfully  following  a  stand-replacement 
fire  in  northwestern  Wyoming.  All  trees  were  killed 
by  the  fire,  which  occurred  55  years  before  this 
photograph  was  taken  in  1 988. 


WHITEBARK  PINE  DECLINE 

Whitebark  pine  has  decHned  in  abundance  in  major  por- 
tions of  its  range  (Arno  1986).  Once  important  on  10  to  15 
percent  of  the  forested  landscape  in  the  Northern  Rocky 
Mountains  (Arno  and  Hoff  1989),  whitebark  pine  mortality 
rates  averaged  42  percent  over  the  last  20  years  in  western 
Montana  (Keane  and  Arno  1993).  Arno  and  others  (1993) 
found  that  for  a  200-ha  study  area  in  Montana,  the  per- 
centage of  stands  with  at  least  20  percent  basal  area  of 
mature  whitebark  pine  declined  from  37  percent  in  1900 
to  20  percent  in  1991;  14  percent  of  the  area  was  dominated 
by  whitebark  pine  in  1900;  none  was  so  dominated  in  1991. 

Whitebark  pine  decline  is  most  pronounced  on  the  more 
productive  sites  where  subalpine  fir  and  Engelmann  spruce 
are  highly  competitive  (Arno  1986;  Ciesla  and  Fumiss  1986; 
Keane  and  Arno  1993;  Kendall  and  Arno  1990).  Declining 
whitebark  pine  threatens  wildlife  habitat  because  the  seeds 
of  whitebark  pine  are  a  valuable  food  for  many  birds  and 
small  mammals,  including  the  endangered  grizzly  bear 
(Kendall  and  Arno  1990). 

Recent  decline  in  whitebark  pine  abundance  is  linked 
to  less  frequent  fires  (Keane  and  Morgan,  these  proceed- 
ings; Keane  and  others  1990).  Fires  in  whitebark  pine  for- 
ests occurred  at  mean  intervals  of  30  to  300  years  based  on 
fire  history  information  derived  from  fire  scars  and  stand 
ages  (table  1).  Fewer  fires  in  the  last  50  to  100  years  (Arno 
and  Hoff  1989;  Morgan  and  Bunting  1990)  have  resulted 
in  extensive  changes  in  the  composition  of  forests  in  the 
high-elevation  landscapes  of  the  Northern  Rocky  Mountains, 
Composition  of  subalpine  forests  has  shifted  dramatically 
toward  dominance  by  subalpine  fir  and  Engelmann  spruce 
(Keane  and  others  1993).  The  decline  of  whitebark  pine 
has  been  further  exacerbated  by  the  introduced  blister  rust 
{Cronartium  ribicola)  and  the  native  mountain  pine  beetle 
(Dendroctonus  ponderosae),  both  of  which  kill  whitebark 
pine  but  not  subalpine  fir  or  Engelmann  spruce. 


STAND-REPLACING  FIRES 

In  conditions  of  extreme  drought  lasting  more  than 
2  years,  fires  ignited  by  lightning  and  fanned  by  high  winds 
can  rapidly  spread  and  kill  trees  in  large  patches.  These 
fires  usually  bum  in  other  forest  types  as  well,  converting 
large  segments  of  the  landscape  to  early  successional  plant 
communities.  Fires  that  spread  through  forests  at  lower 
elevations  historically  burned  into  the  adjacent  whitebark 
pine  forests  (Arno  and  Hoff  1989).  In  whitebark  pine  for- 
ests, stand-replacing  fires  typically  spread  on  the  ground 
(Lasko  1990).  Fires  may  kill  trees  by  scorching  foliage  or 
by  heating  the  bole  or  roots  to  lethal  temperatures.  Some- 
times, crown  fires  occiu-  that  bum  through  the  tree  crowns, 
killing  all  trees  in  their  paths  (Lasko  1990). 

Stand-replacing  fires  provide  important  opportunities 
for  whitebark  pine  to  regenerate.  Many  competing  tree 
species  rely  on  the  wind  to  disseminate  seed.  Whitebark 
pine  has  a  distinct  advantage  in  regenerating  following 
extensive  disturbances  (Tomback  and  others  1990).  The 
Clark's  nutcracker  {Nucifraga  columbiana)  conmionly  trans- 
ports seeds  several  kilometers  (Hutchins  and  Lanner  1982). 
These  birds  prefer  open,  burned  areas  for  caching  seeds 
(Tomback  and  others  1990).  Thus,  although  large  fires  are 
infrequent,  they  are  ecologically  important  in  maintaining 
extensive  whitebark  pine  forests  on  the  landscape. 

LOW-INTENSITY  SURFACE  FIRES 

Low-intensity  surface  fires  also  influence  the  relative 
abundance  of  whitebark  pine  on  the  landscape.  Such 
fires  are  more  frequent  and  smaller  in  extent  than  stand- 
replacing  fires.  Low-intensity  fires  generally  kill  young 
whitebark  pine  and  both  large  and  small  subalpine  fir. 
Such  fires  can  result  in  open,  parklike  stands  of  nearly 
pure  whitebark  pine  (Arno  1986).  Some  fires  probably 
burned  as  low-intensity  surface  fires  but  later  became 
stand-replacing  fires  when  burning  conditions  were  more 
severe.  That  whitebark  pine  trees  often  survive  surface 
fires  is  evidenced  by  the  many  living  trees  that  have  scars 
from  one  or  more  fires  (fig.  1). 


Table  1— Fire  frequency  from  whitebark  pine  forests  expressed 
as  the  mean  and  range  (in  parentheses)  of  the  years 
between  fires 


Fire  frequency 

Geographical  area  and  reference 

144  (55  to  304) 

Bob  Marshall  Wilderness  Complex,  north- 

western fy/lontana  (Keane  and  others  1 993) 

80  (50  to  300) 

Bitterroot  Mountains,  Montana  (Arno  1980) 

30  to  41  (4  to  78) 

1 00-  to  300-ha  stands  where  subalpine  fir 

is  climax,  Montana  (Arno  1986) 

29  (13  to  46) 

10  stands  within  100  ha,  northwestern 

Wyoming  (Morgan  and  Bunting  1990) 

300 

Lodgepole  pine  forests  adjacent  to  but 

at  lower  elevations  than  whitebark  pine. 

Yellowstone  National  Park  (Romme  1 982) 

138 


These  low-intensity  fires  are  more  common  on  relatively 
dry  sites,  occurring  only  where  stand  structures,  fuel  ac- 
cumulation, and  microclimatic  conditions  are  conducive. 
Thus,  such  fires  result  in  many  small  burned  patches,  in- 
creasing landscape  heterogeneity. 

Where  fires  are  more  fi'equent,  they  are  more  likely  to  be 
of  low  intensity.  Morgan  and  Bimting  (1990)  documented 
very  fi'equent  low-intensity  fires  on  a  relatively  dry  site 
supporting  serai  whitebark  pine  in  open,  parkHke  stands 
within  a  100-ha  area  in  northwestern  Wyoming.  There  the 
mean  interval  between  fires  was  33  years  prior  to  1867. 
Fires  were  much  more  common  prior  to  1850  than  they 
have  been  since  then  (Morgan  and  Bunting  1990). 

FIRE  REGIMES 

Through  time,  most  whitebark  pine  forests  experience  a 
mixture  of  stand-replacement  and  low-intensity  fires.  The 
frequency  of  these  types  of  fires  within  a  given  landscape 
will  vary  with  landscape  complexity  and  heterogeneity. 
Stand-replacing  fires  are  more  common  diiring  regional 
droughts,  and  often  burn  large  patches  regardless  of  fuel 
loading  or  stand  condition.  Fire  behavior  and  effects  are 
also  influenced  by  the  stand  structure  and  fuel  accumula- 
tion, which  are  in  part  determined  by  the  time  since  last 
bum. 

Where  whitebark  pine  is  cHmax,  fires  are  infrequent  and 
generally  of  low  intensity.  In  whitebark  pine  krummholz 
and  ribbon  forests,  fires  are  infrequent  and  of  variable  in- 
tensity. When  fires  do  occ\ir,  many  trees  die  and  regenera- 
tion is  very  slow.  Keane  and  others  (1990)  predict  where 
blister  rust  infection  rates  are  high,  climax  whitebark  pine 
forests  will  convert  to  herbaceous  or  shrub  commtmities 
following  fire. 

Stand-replacing  fires  are  more  common  where  whitebark 
pine  is  a  serai  dominant.  Stand-replacing  fires  become  in- 
creasingly likely  with  advancing  succession  (Fischer  and 
Clayton  1983;  Morgan  and  Bimting  1990). 

SUCCESSION  FOLLOWING  FIRE 

Fire  is  a  key  process  affecting  serai  whitebark  pine  forest 
structure  sind  composition.  Successional  patterns  on  sites 
where  whitebark  pine  is  serai  are  predictable  (fig.  5),  but 
they  are  not  closely  tied  to  stand  age  or  time  since  last  dis- 
tiu-bance  (Mattson  and  Reinhart  1990).  The  stand  struc- 
ture and  the  microsites  created  vary  from  fire  to  fire.  Con- 
ditions for  successful  regeneration  of  tree  seedlings  are 
sporadic,  depending  on  favorable  climatic  and  site  condi- 
tions. Althoiigh  both  subalpine  fir  and  whitebark  pine 
may  establish  soon  after  a  fire,  it  may  take  a  half-century 
or  longer  for  a  forest  to  develop. 

Whitebark  pine  is  one  of  the  first  tree  species  to  become 
established  in  abimdance  following  stand-replacing  fires 
(Weaver  and  Dale  1974).  As  a  consequence,  it  often  domi- 
nates initially,  often  for  up  to  225  years  or  more  (Loope 
and  Gruell  1973;  Morgan  and  Bimting  1990).  Early  serai 
stands  are  dominated  by  whitebark  pine  seedlings  and  sap- 
lings growing  along  with  a  dense  herbaceous  and  shrub 
understory.  Subalpine  fir  seedlings  are  often  present, 
especially  close  to  parent  trees  that  survived  the  fire,  but 
they  grow  more  slowly  than  whitebark  pine  trees  (Arno 


Late  Serai 
(ABLA,  PEN, 
PIAL,  PICO) 


Shrub/herb 


Late  Mid-Seral 
(ABLA,  PIAL, 
PICO,  PEN) 


Early  Sera! 
(PIAL,  PICO 
ABLA,  PEN) 


Mid-Serai 
(PIAL,  PICO, 
ABLA,  PEN) 


^^^^^^ 


Stand-replacing  fire 
Return  to  shrub/herb  stage 


Low-intensity  fire 
Many  PIAL  survive 


Figure  5 — Generalized  forest  succession  following 
fires  on  sites  where  whitebark  pine  (PIAL)  is  re- 
placed by  subalpine  fir  (ABLA)  and  Engelmann 
spruce  (PEN)  with  advancing  succession.  Lodge- 
pole  pine  (PICO)  is  a  common  associate.  Low- 
intensity  surface  fires  may  occur  at  any  stage 
but  are  most  likely  in  mid-seral  stands.  Stand- 
replacing  fires  are  the  norm  in  late-seral  stands. 
Tree  species  listed  in  order  of  abundance. 
Adapted  from  Fischer  and  Clayton  (1983). 


and  Hoff  1989).  Whitebark  pine  seedlings  are  more  abun- 
dant than  subalpine  fir  seedlings  in  large  burned  areas 
because  whitebark  pine  seedlings  are  dispersed  farther  by 
the  Clark's  nutcracker.  Subalpine  fir  seeds  are  dispersed 
by  the  wind.  Many  standing  snags  and  fallen  logs  are 
present.  Most  whitebark  pine  trees  do  not  produce  large 
numbers  of  seeds  imtil  at  least  age  70;  most  do  not  produce 
any  cones  xmtil  age  50  (Morgan  and  Bvmting  1992).  With 
time,  whitebark  pine  is  gradually  replaced  by  subalpine  fir 
and  Engelmann  spruce.  In  late-seral  stands,  both  the  tree 
canopy  and  the  imderstory  are  dominated  by  many  subal- 
pine fir  trees. 

MANAGEMENT  IMPLICATIONS 

Continued  decline  in  whitebark  pine  abundance  threat- 
ens to  dramatically  reduce  the  availability  of  seeds  for  the 
many  animals  that  rely  on  them  as  a  food  source  (Arno 
1986). 

Fire  exclusion  greatly  reduces  opportimities  for  regen- 
eration of  whitebark  pine.  Cone  production  is  higher  in 
the  stands  where  whitebark  pine  is  healthy  and  dominant 
(Morgan  and  Bunting  1992).  With  blister  rust  reducing 
cone  production  and  killing  parent  trees  (Arno  and  Hoff 
1989),  the  seed  available  for  tree  regeneration  is  rapidly 


139 


declining  (Amo  1986;  Keane  and  Amo  1993).  As  cone  pro- 
duction declines,  animals  eat  more  of  the  seeds,  leaving 
fewer  to  regenerate.  Whitebark  pine  cone  production  de- 
clines with  advancing  succession  (Morgan  and  Bunting 
1992)  and  as  infection  by  blister  rust  increases  (Arno  and 
Hoff  1989).  Given  the  rapidly  declining  abundance  of  white- 
bark  pine  in  some  regions,  we  must  act  quickly  to  create  op- 
portunities for  individuals  that  are  resistant  to  blister  rust 
to  regenerate.  Although  whitebark  pine  is  very  sensitive 
to  blister  rust,  some  individual  trees  are  genetically  resis- 
tant to  infection  (Hoff  and  others  1990).  If  we  do  not  create 
opportunities  for  those  trees  to  regenerate  before  their  cone 
production  declines  or  they  die  in  advancing  succession, 
opportunities  for  enhancing  natural  mechanisms  of  white- 
bark pine  recovery  will  be  lost. 

Three  options  are  available  for  improving  the  health  and 
productivity  of  whitebark  pine  stands  that  are  now  domi- 
nated by  subalpine  fir  and  spruce.  One  option  is  creating 
forest  openings  through  timber  harvest,  girdling,  or  other- 
wise killing  trees  mechanically.  Cutting  trees  to  create 
openings  to  encourage  regeneration  of  whitebark  pine  is 
possible  but  may  not  be  economically  feasible.  Although 
there  are  often  small  whitebark  pine  trees  growing  in  the 
imderstory  of  mixed  conifer  stands,  they  are  often  very  old 
and  of  poor  vigor  and  are  therefore  unlikely  to  respond 
when  larger  trees  are  removed. 

Another  option  is  the  liberal  use  of  prescribed  fire  to  cre- 
ate openings  for  regeneration  and  to  favor  whitebark  pine 
in  stands  now  codominated  by  subalpine  fir.  Managers  can 
purposefully  ignite  fires  or  allow  lightning  fires  to  burn 
under  carefully  prescribed  conditions  of  weather,  fuel,  and 
location.  Historically,  fire  was  the  primary  natural  distur- 
bance, and  it  may  be  the  most  practical  tool  for  managing 
whitebark  pine  considering  economics,  policy,  and  topo- 
graphic limitations  for  using  timber  harvest. 

A  third  option,  to  use  a  combination  of  techniques,  may 
be  most  successful.  Arno  and  Keane  are  involved  in  an  ef- 
fort to  test  alternative  techniques  for  perpetuating  white- 
bark pine  in  a  blister  rust-infected  area  of  the  Bitterroot 
National  Forest  in  Montana. 

ACKNOWLEDGMENTS 

Research  funding  was  provided  through  the  USDA  Forest 
Service,  Intermountain  Research  Station.  This  is  Contribu- 
tion No.  693  of  the  Idaho  Forest,  Wildlife  and  Range  Experi- 
ment Station,  University  of  Idaho,  Moscow,  ID  83843. 
We  thank  Bill  Fischer  and  Elizabeth  Reinhardt  for  their 
comments. 

REFERENCES 

Arno,  S.  1980.  Forest  fire  history  in  the  Northern  Rockies. 
Journal  of  Forestry.  78(8):  460-465. 

Arno,  S.  1986.  Whitebark  pine  cone  crops — a  diminishing 
source  of  wildlife  food?  Western  Journal  of  Applied  For- 
estry. 1(3):  92-94. 

Arno,  S.  F.;  Hoff,  R.  J.  1989.  Silvics  of  whitebark  pine 
{Pinus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  lip. 


Arno,  S.  F.;  Reinhardt,  E.  D.;  Scott,  J.  H.  1993.  Forest 
structure  and  landscape  patterns  in  the  subalpine  lodge- 
pole  pine  t)^es:  a  procedure  for  quantifying  past  and 
present  conditions.  Gen.  Tech.  Rep.  INT-294.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service, 
Intermountain  Research  Station.  17  p. 

Ciesla,  W.  M.;  Furniss,  M.  M.  1986.  Idaho's  haimted  forests. 
American  Forests.  81(8):  23-25. 

Fischer,  W.  C;  Clayton,  B.  D.  1983.  Fire  ecology  of  Montana 
forest  habitat  types  east  of  the  Continental  Divide.  Gen. 
Tech.  Rep.  INT-141.  Ogden,  UT:  U.S.  Department  of 
Agriculture,  Forest  Service,  Intermountain  Forest  and 
Range  Experiment  Station.  83  p. 

Hoff,  R.  J.;  Bingham,  R.  T.;  McDonald,  G.  I.  1980.  Relative 
blister  rust  resistance  of  white  pines.  European  Journal 
of  Forest  Pathology.  10:  307-316. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Keane,  R.  E.;  Arno,  S.  F.  1993.  Rapid  decline  of  whitebark 
pine  in  western  Montana:  evidence  from  10-year  meas- 
urements. Western  Journal  of  Applied  Forestry.  8(2): 
44-47. 

Keane,  R.  E.;  Arno,  S.  F.;  Brown,  J.  K;  Tomback,  D.  F. 
1990.  Modelling  stand  dynamics  in  whitebark  pine 
forests.  Ecological  Modelling.  51:  73-95. 

Keane,  R.  E.;  Morgan,  P.;  Menakis,  J.  1993.  Landscape 
assessment  of  the  decline  of  Pinus  albicaulis  in  the 
Bob  Marshall  Wilderness  Complex,  Montana,  USA. 
Missoula,  MT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station,  Intermountain 
Fire  Sciences  Laboratory.  Review  draft. 

Kendall,  K  C;  Arno,  S.  F.  1990.  Whitebark  pine — an  impor- 
tant but  endangered  wildlife  resource.  In:  Schmidt,  W.  C; 
McDonald,  K.  J.,  comps.  Proceedings — sjmaposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  264-273. 

Lasko,  Richard  J.  1990.  Fire  behavior  characteristics  and 
management  implications  in  whitebark  pine  ecosystems. 
In:  Schmidt,  W.  C.;  McDonald,  K.  J.,  comps.  Proceedings — 
sjmiposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station:  319-323. 

Loope,  L.  L.;  Gruell,  G.  E.  1973.  The  ecological  role  of  fire 
in  the  Jackson  Hole  area,  northwestern  Wyoming.  Qua- 
ternary Research.  3:  425-443. 

Mattson,  D.  J.;  Reinhart,  D.  P.  1990.  Whitebark  pine  on  the 
Mount  Washburn  Massif,  Yellowstone  National  Park.  In: 
Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  Proceedings- 
symposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain Research  Station:  106-117. 

Morgan,  P.;  Bunting,  S.  C.  1990.  Fire  effects  in  whitebark 
pine  forests.  In:  Schmidt,  W.  C;  McDonald,  K.  J.,  comps. 


140 


Proceedings — symposium  on  whitebark  pine  ecosystems: 
ecology  and  management  of  a  high-momitain  resource; 
1989  March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep. 
INT-270.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermoimtain  Research  Station:  166-170. 

Morgan,  P.;  Bunting,  S.  C.  1992.  Cone  crops  in  whitebark 
pine  (Pinus  albicaulis)  forests  in  the  Greater  Yellowstone 
Area.  Moscow,  ID:  University  of  Idaho,  College  of  For- 
estry, Wildlife  and  Range  Sciences.  21  p.  Review  draft. 

Pfister,  R.  D.;  Kovalchik,  B.  L.;  Arno,  S.  F.;  Presby,  R.  C. 
1977.  Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Experi- 
ment Station.  174  p. 

Pyne,  S.  J.  1982.  Fire  in  America.  Princeton,  NJ:  Princeton 
University  Press.  654  p. 

Romme,  W.  H.  1982.  Fire  and  landscape  diversity  in  subal- 
pine  forests  of  Yellowstone  National  Park.  Ecological 
Monographs.  52:  199-221. 

Steele,  R.;  Pfister,  R.  D.;  Ryker,  R.  A.;  Kittams,  J.  A.  1981. 
Forest  habitat  types  of  central  Idaho.  Gen.  Tech.  Rep. 


INT-114.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  138  p. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.; 

Pfister,  R.  D.  1983.  Forest  habitat  types  of  eastern  Idaho- 
western  Wyoming.  Gen.  Tech.  Rep.  INT- 144.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Forest  and  Range  Experiment  Station.  122  p. 

Tomback,  D.  F.;  Hoffman,  L.  A.;  Sund,  S.  K.  1990.  Coevo- 
lution  of  whitebark  pine  and  nutcrackers:  implications 
for  forest  regeneration.  In:  Schmidt,  W.  C;  McDonald, 
K.  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermountain  Research 
Station:  118-130. 

Weaver,  T.;  Dale,  D.  1974.  Pinus  albicaulis  in  central 
Montana:  environment,  vegetation  and  production. 
American  Midland  NaturaHst.  92:  222-230. 


141 


VEGETATION  DISTRIBUTION  AND 
PRODUCTION  IN  ROCKY  MOUNTAIN 
CLIMATES— WITH  EMPHASIS  ON 
WHITEBARK  PINE 

Tad  Weaver 


Abstract — The  distribution  and  production  of  vegetation  on  the 
altitudinal  gradient  (grassland-forest-alpine)  was  plotted  against 
climatic  parameters  to  evaluate  hypothetical  controlling  factors. 
(1)  Whitebark  pine  (Pinus  albicaulis)  is  likely  excluded  from 
higher  zones  by  a  cool  growing  season  or  wind-induced  drought. 
It  is  probably  not  excluded  by  low  temperatures  occurring  during 
its  hardening,  hard,  or  dehardening  seasons.  (2)  While  the  lower 
physiological  limit  of  whitebark  pine  is  probably  set  by  drought, 
its  lower  realized  limit  is  directly  set  by  subalpine  fir  (Abies  lasio- 
carpa)  and  lodgepole  pine  (Pinus  contorta)  competitors  and  indi- 
rectly set  by  factors  that  control  their  distribution.  (3)  The  upper 
limits  for  most  other  dominant  species  are  probably  set  by  grow- 
ing season  temperature.  The  lower  Umits  are  likely  set  by  com- 
petition down  to  the  cedar-hemlock  (Thuja  plicata  I  Tsuga  hetero- 
phylld)  zone  and  by  drought  in  drier  areas.  (4)  Production  is 
strongly  correlated  (r^  =  0.86)  with  growing  season  length  (soil 
thawed  season  minus  dry  soil  days).  Multiplying  season  length 
by  average  temperature  did  not  improve  the  growing  season  pre- 
dictor, perhaps  because  vegetation  at  each  altitude  is  especially 
adapted  to  temperatures  in  its  zone. 


Vegetation  composition  and  structure  vary  along  gradi- 
ents of  temperature  and  precipitation  v^^hether  the  condi- 
tion changes  with  altitude  (Daubenmire  1956;  Gams  1931; 
Weaver  1980)  or  geography  (Holdridge  1967;  Walter  1973; 
Whittaker  1975).  Graphical  devices  based  on  simple  pa- 
rameters such  as  average  annual  temperature  and  total 
annual  precipitation  (Holdridge  1967;  Whittaker  1975) 
are  workable  predictors  of  the  vegetation  growing  in  par- 
ticular climates.  It  seems  obvious,  however,  that  these 
devices  succeed,  not  because  the  climatic  parameters  used 
are  causal,  but  because  the  parameters  are  correlated 
with  causal  climatic  factors. 

Production  of  vegetation  is  largely  determined  by  cli- 
mate and  is  therefore  expected  to  be  highest  where  tem- 
perature, moisture,  and  nutrients  are  simidtaneously 
favorable.  In  Rocky  Mountain  vegetation  one  therefore 
expects  production  to  be  highest  in  low-altitude  (warm) 
moist  forests  (Thuja-Tsuga)  and  to  decline  both  with  in- 
creasing altitude  (because  of  cooling  and  reduced  nutri- 
ent availability)  (Weaver  1979)  and  with  decreasing  alti- 
tude (because  of  decreasing  water  availability).  While  it  is 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Tad  Weaver  is  Plant  Ecologist,  Biology  Department,  Montana  State 
University,  Bozeman,  MT  59717. 


rarely  done,  vegetation  production  should  also  be  success- 
fully plotted  in  a  Hutchinsonian  (1958)  hyperspace  with 
axes  that  are  either  physiologically  meaningful  or  surro- 
gates well  correlated  with  physiologically  meaningful  axes. 

The  object  of  this  paper  is  to  correlate  vegetation 
performance — survival  and  production — ^with  physiologi- 
cally meaningftd  aspects  of  temperature  and  moisture. 
Study  of  the  distribution  of  species  on  these  prestunptively 
causal  axes  will  eliminate  some  hypotheses  of  cause  and 
sharpen  others  for  experimental  tests.  While  whitebark 
pine  (Pinus  albicaulis)  is  the  primary  subject  of  this  paper, 
the  approach  could  be  applied  to  other  species,  as  well. 

METHODS 

Impacts  of  climatic  factors  on  vegetation  distribution 
were  evaluated  by  comparing  factor  levels  among  seg- 
ments on  a  vegetational  gradient.  In  essence,  the  method 
allows  one  to  deduce  (1)  that  a  presmnptive  factor  that 
does  not  vary  between  vegetation  t3T)es  is  not  controlling 
the  vegetation  changes  observed,  and  (2)  that  a  presimip- 
tive  factor  that  does  change  between  vegetation  types  de- 
serves further  discussion  as  directly  controlhng,  indirectly 
controlling,  or  a  noncausal  correlate.  The  method  is  illus- 
trated here  with  discussion  of  whitebark  pine,  but  it  could 
be  applied  to  many  other  species  as  well. 

The  Data 

Environmental  zones  were  identified  by  climax  vegeta- 
tion occupying  them  (Daubenmire  1943;  Daubenmire  and 
Daubenmire  1968;  Huschle  and  Hironaka  1980).  They  in- 
cluded, from  low  to  high  altitude,  desert  shrub,  dry  grass- 
land, warm  forest,  cool  forest,  and  alpine  ecosystems. 
Specific  zones  are  listed  with  their  Kuchler  (1964)  type 
numbers  and  representative  stations  in  table  1. 

Climates  of  the  environmental  zones  were  characterized 
by  using  climatic  statistics  from  approximately  five  sta- 
tions in  each  environmental-vegetation  zone.  To  maintain 
the  integrity  of  the  data,  raw  data  are  plotted  wherever 
possible.  When  necessary,  medians  were  used  to  repre- 
sent "typical"  sites;  use  of  medians  deemphasizes  sites 
vrith  conditions  intermediate  between  modal  conditions 
of  adjacent  types.  The  stations  chosen  include  most  avail- 
able stations  and  may  thus  be  considered  as  a  "complete 
sample."  The  sample  may  not  be  entirely  random  because 
weather  stations  must  be  accessible;  in  higher  types,  for 
example,  the  sites  may  be  relatively  low. 


142 


Table  1 — Weather  stations  representative  of  major  Northern  Rocky  Mountain  environmental  zones. ^  Zones  are  listed  from  high  to  low  altitude 


Zone  Station  locations 


Alpine^ 

Whitebark  pine^ 
Subalpine  fir 
Douglas-fir 
Cedar-hemlock 
Ponderosa  pine 

Idaho  fescue 
Bluebunch 
wheatgrass 
Grama  grass 
Atriplex  spp. 


White  Mountain,  CA;  Niwot  Ridge,  CO 

Old  Glory,  BC;  Kings  Hill,  MT;  Ellery  Lake,  CA;  Crater  Lake,  OR 

High:  Summit,  MT;  West  Yellowstone,  MT;  Lake  Yellowstone,  MT;  Low:  Burke,  ID;  Seeley  Lake,  MT;  Hungry  Horse,  MT 
East:  Hebgen  Lake,  MT;  Lakeview,  MT;  Lamar,  WY;  Palisades  Dam,  ID;  Dixie,  ID;  West:  Libby,  MT;  Lincoln,  MT 
West  Glacier,  MT;  Sandpoint,  ID;  Pierce,  ID;  Avery,  ID;  Priest,  ID 

West:  Garden  Valley,  ID;  Poison,  MT;  Potlatch,  ID;  Kooskia,  ID;  East:  Melstone,  MT;  Busby,  MT;  Colstrip,  MT;  Lame 
Deer,  MT;  Roundup,  MT 

Gallatin  Gateway,  MT;  Virginia  City,  MT;  Mystic  Lake,  MT;  Wisdom,  MT;  Bozeman  MSU,  MT;  White  Sulphur  Springs,  MT 

Browning,  MT;  Belgrade,  MT;  Kalispell,  MT;  Ennis,  MT;  Dillon,  MT;  Augusta,  MT;  Crow  Agency,  MT;  Billings,  MT 
Ekalaka,  MT;  Jordan,  MT;  Malta,  MT;  Rock  Springs,  MT;  Rapelje,  MT;  Chester,  MT;  Great  Falls,  MT;  Fairfield,  MT 
Lovell,  WY;  Worland,  WY;  Basin,  WY;  Powell,  WY;  Deaver,  WY 


'The  vegetation  types,  with  their  Kuchler  (1964)  type  numbers  (KTXX),  are:  Deschampsia  caespitosa-Carex  (KT52),  Pinus  albicaulis  (KT15),  Abies  lasiocarpa 
(KT1 5),  Pseudotsuga  menziesii  (KT1 2),  Thuja  plicataJTsuga  heterophylla  (KT2),  Pinus  ponderosa  (KT1 1,16),  Festuca  idatioensis  (KT63),  Agropyron  spicatum 
(KT63),  Bouteloua  gracilis  (KT64),  Atripiex  (KT40).  Abies  types  from  higher  and  lower  altitudes,  Pseudotsuga  types  from  east  and  west  of  the  mountains,  and  pon- 
derosa pine  types  from  east  and  west  of  the  Rockies  are  separately  presented  to  emphasize  possible  differences. 

n"he  White  Mountain  alpine  site  (1 0,1 50  ft)  is  unusually  dry.  The  Ellery  Lake  whitebark  pine  site  has  only  precipitation  data;  temperature  data  were  therefore 
taken  from  a  similar  altitude  at  a  nearby  White  Mountain  site  (9,645  ft). 


Killing  Temperatures 

The  hypothesis  that  extreme  temperatiires  might  ex- 
clude a  plant  (for  example,  whitebark  pine)  from  an  adja- 
cent zone  was  tested  by  determining  whether  the  condi- 
tion is  more  extreme  in  the  iinoccupied  than  in  the 
occupied  zone;  if  it  is  not,  the  hypothesized  factor  is  con- 
sidered unlikely  to  be  lethal.  This  test  was  used  on  tem- 
peratures of  early  fall  frosts,  midwinter  extreme  lows, 
midwinter  average  lows,  late  fall  frosts,  midsummer  aver- 
age highs,  and  midsummer  extreme  highs.  Midwinter 
lows  £ind  midsvmimer  highs  were  tested  because  these 
temperatures  might  exceed  the  potential  tolerances  of  a 
plant;  fall  and  spring  frosts  were  considered  because  they 
might  kill  partially  hardened  plants.  Fall  frost  tempera- 
tiu-es  were  measured  as  the  absolute  low  in  the  month  in 
which  average  temperature  first  fell  below  0  °C;  where  no 
month  had  an  average  temperatvire  of  0  °C,  the  values 
were  interpolated  to  0  °C  from  adjacent  months.  Spring 
frost  temperatiu-es  were  estimated  similarly  for  the 
month  in  which  average  temperatures  first  rose  above 
0  °C.  I  argue  later  (see  growing  season)  that  average  tem- 
peratures below  0  °C  restrict  water  and  nutrient  uptake 
and  therefore  open  and  close  the  growing  season;  the  use 
of  0  °C  is  in  contrast  to  the  5  °C  used  by  many  phenolo- 
gists  (for  example,  Chang  1968).  "Absolute"  temperatures 
were  the  lowest  (or  highest)  seen  in  the  period  of  record — 
10  years  or  more. 

Starvation  Temperatiu*es 

The  plant  might  be  excluded  from  a  zone  free  from  kill- 
ing events  if  (temperatiire)  conditions  failed  to  support 
net  photosynthesis  on  an  annual  basis.  Two  indices  of 
growing  season  temperature  were  used  to  test  the  possi- 
bility of  such  starvation.  First,  temperatures  were  aver- 
aged across  all  growing  season  (defined  later)  months. 
Despite  its  common  use  in  predicting  growth  (Chang  1968; 
Larcher  1975),  this  index  is  expected  to  underestimate  the 


benefits  of  warmer  temperatvires  because  chemical  reaction 
rates  increase  exponentially  with  increasing  temperature. 
An  alternate  index  of  temperature  was  therefore  tested. 
In  the  second  index  temperatures  were  replaced  with 
growth  support  units  (gsu)  and  these  were  averaged 
across  the  season.  On  the  assimiptions  that  Q^^  =  2,  that 
very  slow  growth  begins  at  1  °C,  and  that  native  plants 
tolerate  normal  high  temperatures  of  their  zones,  the 
growth  support  unit  curve  was  constructed  by  interpolat- 
ing between  0  growth  rate  imits  (gsu)  at  0  °C,  1  gsu  at 
1  °C,  2  gsu  at  11  °C,  4  gsu  at  21  °C,  8  gsu  at  31  °C,  and 
16  gsu  at  41  °C.  That  is,  we  expect  no  growth  at  0  °C  or 
below,  1  vmit  at  1  °C,  2  imits  at  11  °C,  4  units  at  21  °C,  etc. 

Killing  Drought 

Evaporation  equals  roughly  2  mm/°C  x  month 
(Daubenmire  1956;  Nielsen  1986;  Stephenson  1990; 
Thornthwaite  1948;  Walter  1973).  As  a  result,  one  might 
index  drought  duration  by  counting  months  when  average 
precipitation  (mm/2)  is  less  than  average  temperature 
(C)  or  drought  intensity  by  summing,  across  growing  sea- 
son months,  precipitation  deficits  below  the  calculated 
balance. 

A  covmt  of  months  registering  drought  on  the  Walter  in- 
dex may  overestimate  drought  duration  where  the  deficit 
is  small  because  drought  may  open  late  or  close  early  in 
the  month.  To  minimize  this  effect  I  have  normalized 
drought  duration  from  a  scatter  diagram  of  duration 
against  deficit:  0-5  mm  =  0  months,  6-15  mm  =  1  month, 
16-35  mm  =  2  months,  36-50  mm  =  3  months,  and 
51-85  mm  =  4  months. 

I  have  used  the  simi  of  warm-season  (average  tempera- 
ture above  0  °C)  water  deficits  as  an  index  of  drought 
magnitude.  I  do  so  with  the  recognition  that,  due  to  ex- 
clusion of  important  factors  (for  example,  wind,  Penman 
1949,  and  exponential  temperature  effects,  Stephenson 
1990),  this  index  usually  underestimates  water  deficit. 


143 


Water  Starvation 


RESULTS  AND  DISCUSSION 


"Water  starvation"  should  occur  where  plants  are  not 
desiccated,  but  where  water  supplies  do  not  support  net 
photosynthesis  on  an  annual  basis.  For  example,  starva- 
tion would  occur  if  water  were  continually  available,  but 
evapotranspiration  equaled  or  exceeded  uptake  rates.  Such 
starvation  would  select  for  organisms  with  lower  ET  rates: 
Mesophytic  leaves  would  be  replaced  by  xeroph)i;ic  leaves 
and  leaf  exposiu-e  might  be  reduced  by  reduction  of  stature 
even  to  the  point  where  plants  existed  only  at  the  groimd 
surface  or  under  transparent  rock. 

Growing  Season 

For  sensitive  introduced  plants,  frost-free  season  is  of- 
ten recorded  as  an  indicator  of  season  length  (Burke  and 
others  1976;  USGS  1971). 

Since  cool  region  plants  normally  tolerate  growing  sea- 
son frosts,  I  argue  for  a  different  index.  Growth  requires 
both  leaf  activity  (photos3nithesis)  and  root  activity 
(water  and  nutrient  absorbtion).  For  plants  with  frost- 
insensitive  leaves  (Burke  and  others  1976),  midday  tem- 
peratures should  be  high  enough  for  photosynthesis  long 
before  soils  are  thawed.  Root  activity  extends  (maxi- 
mally) from  spring  soil  thaw  to  the  fall  refreeze,  so  this 
period  should  be  a  better  index  of  growing  season  length. 
Thus  I  open  and  close  the  growing  season  with  0  °C  aver- 
age monthly  air  temperatures.  So  represented,  initial  ac- 
tivity in  grasslands  begins  at  spring  soil  temperatures 
(0-25  cm)  near  0  °C  and  activity  ceases  in  the  fall  with  soil 
temperatures  as  high  as  5  °C  (Weaver,  in  preparation). 
In,  or  adjacent  to,  the  subalpine  fir  {Abies  lasiocarpa)  zone 
the  growing  season  may  exceed  this  period  because  deep 
snow  cover  may  prevent  the  freezing  of  soil  water. 

While  the  temperature-bounded  season  probably  ap- 
plies at  high-altitude  sites,  at  low-altitude  sites  soil  dry- 
ing may  close  the  growing  season  before  frost  does.  Thus, 
one  may  better  index  growing  season  as  warm  season 
months  minus  any  drought  months  included  in  the  warm 
season.  While  the  Northern  Rocky  Mountain  drought  sea- 
son comes  at  summer's  end  (Weaver  1980),  this  index  will 
also  apply  to  regions  where  the  warm  season  opens  with 
dry  months.  It  is  conceivable  that  other  periodic  factors, 
for  example  nutrients  or  pest  attack,  might  affect  growing 
season  length. 

Production 

Production  estimates  for  each  series  were  drawn  from 
the  literature  (table  2).  These  are  correlated  with  climatic 
conditions  that  might  control  them:  warm  season  length, 
warm  moist  season  length,  and  the  product  of  warm  soil 
season  length  and  temperature.  For  the  latter,  two  tem- 
perature expressions  were  tested:  average  temperature 
minus  5  °C  and  Q  (the  Q^^-based  average).  The  third 
(product)  indices  were  expected  to  be  best  because  they 
assume  that  plants  grow  when  soils  are  unfrozen-moist 
and  that  their  growth  rate  is  proportional  to  temperature 
(over  5  °C  for  the  first  linear  index  and  over  1  °C  for  the 
second  exponential  index). 


In  the  Northern  Rocky  Motmtains  vegetation  changes 
from  grass-shrub  (Atriplex,  Bouteloua,  Agropyron,  and 
Festuca)  to  conifer  {Pinus  ponderosa,  Thuja-Tsuga, 
Pseudotsuga,  and  Abies)  to  alpine  graminoid  with  increas- 
ing altitude.  Subzones  listed  parenthetically  are  described 
by  Daubenmire  (1943)  and  mapped  by  Kuchler  (1964). 
This  paper  compsu-es  weather  data  gathered  in  these  zones 
to  test  h3TJothesized  controls  of  distribution  and  production. 
The  method  is  illustrated  with  whitebark  pine. 

Upward  Limits 

I  speculate  that  the  upward  limits  of  whitebark  pine 
might  be  determined  by  lethal  factors  (winter  low  tem- 
peratures, spring-fall  fi*osts,  or  desiccation)  or  production 
deficits  (starvation)  due  to  inadequate  water  or  heat 
units.  These  h3T)otheses  are  considered  in  the  following 
paragraphs. 

If  winter  lows  are  lower  in  the  alpine  than  in  the  pine 
zone  below,  winter  cold  might  be  the  excluding  factor,  oth- 
erwise not.  Contrary  to  my  expectation,  neither  absolute 
lows  nor  average  January  minima  are  lower  in  the  alpine 
than  in  the  pine  (or  most  other)  vegetation  zone(s)  (fig.  1). 
The  failure  of  low  temperatures  to  develop  at  higher  alti- 
tudes may  be  due  to  the  high  density  of  cold  air:  Cold  air 
entering  from  the  north  stays  low  like  mercury  poured  \m- 
der  water  and  air  cooled  by  exposure  to  cold  alpine  ground 
runs  off  (Geiger  1965).  Since  extreme  temperatures  are 
not  lower  at  high  than  low  altitude,  I  deduce  that  neither 
whitebark  pine,  nor  most  other  native  dominants,  are  ex- 
cluded from  high  sites  by  extreme  low  temperatures  of  win- 
ter. An  exception  to  this  generalization  seems  to  appear  in 
high-  or  low-altitude  fi-ost  pockets,  where  accumulating 
cold  air  may  kill  trees  in  winter  or  spring  (Weaver  1990). 

For  natives  of  an  area,  I  suggest  that  frost  damage  is 
more  likely  when  plants  are  partially  hardened  (fall)  or 
partially  dehardened  (spring)  than  at  midwinter.  I  expect 
(1)  winter  to  be  delimited  approximately  by  the  fall  month 
in  which  surface  soils  fall  below  0  °C  and  the  spring 
month  in  which  they  rise  above  0  °C,  (2)  winter  to  be  a 
season  of  low  root  activity,  low  water,  and  low  nutrient 
uptake,  and  (3)  0  °C  soil  temperatures  to  be  approxi- 
mately coincident  with  0  °C  air  temperatures  (Weaver,  in 
preparation).  I  therefore  (1)  compare,  across  vegetation 
types,  observed  temperature  lows  in  these  spring-fall  sea- 
sons of  incomplete  frost  hardness  (fig.  2),  (2)  discover  no 
difference,  and  (3)  conclude  that  frosts  probably  do  not 
partition  vegetation  native  to  the  region.  The  preceding 
discussion  depends  on  the  assumption  that  plants  harden 
and  deharden  in  phase  with  air/soil  temperatures.  This 
argument  would  be  fallacious  if  day  length  were  the  pri- 
mary controller  of  the  hardening/dehardening  process. 
However,  since  both  temperature  and  correlated  day 
length  triggers  are  important  (Salisbury  and  Ross  1992), 
I  expect  frosts  to  exclude  plants  from  environmentally  dis- 
tant, but  not  environmentally  adjacent,  vegetation  types. 

While  frosts  and  winter  freezes  probably  do  not  exclude 
whitebark  pine  from  the  alpine  (or  other  plants  from  the 
vegetation  zone  immediately  above),  a  lack  of  heat  imits 


144 


Table  2 — Productivity  in  grams  per  square  meter  per  year  and  aboveground  standing  crop  (Abvegr  std  crp)  of  major  Rocky  Mountain  vegetation 
types 


Vegetation  type 

Productivity^ 

AKuonr  QtH  f^m 

g/m^/yr 

t/ha  (age.yr) 

Alpine 

135 

1.35 

Thilenlus  and  others  1974 

100-200 

1.00-2.00 

Scott  and  Billings  1 964 

Pinus  albicaulis 

'60 

^140  ^300-500^ 

VVCdVd   Cll  IVJ  L^ClIC'    1      /  H 

'25-75  (53) 

PfiQtpr  anH  nthorQ  1Q77 

r  IIOLC7I    al  IVJ  \JlllC7IO    1*7/  1 

200-700 

350  (300-500) 

Forcella  and  Weaver  1 977 

Abies  lasiocarpa 

1 00-200 

160  f300-700) 

Aoipt  and  other*;  1QfiQ 

860 

357  M06  vr) 

Whitt^^kpr  ^r\c\  Niprinn  1  07*^ 

V  V  1  II  llCir\C7l  ai  \\a  I  NIOI  ll  im   I  ^  r  w 

'95-180  (137) 

Pfister  and  others  1 977 

100-250  (Old) 

Weaver  and  Forcella  1977 

150-250 

Landis  and  Mogren  1975 

Ps6udotsuga  menziesii 

1,550 

438  (252) 

Whittakpr  and  Niprinn  1Q7^ 

'30-170  (150) 

Pfi^tpr  and  nthpr^  1Q77 

100-350  (350+) 

Weaver  and  Forcella  1 977 

Thuja  plicata- 

Tsuaa  hsteroohvlla 

870 

Hanlfiv  1  Q7fi 
rial  Moy  i  7 /  u 

550 

316  (250) 

Hanlev  1976 

1,380 

504  (103) 

Hanley  1976 

'150-330  (240) 

Pfister  and  others  1 977 

Pinus  pondorosa 

'188 

(^larpv/  ^inH  Athorc  1Q7^ 

'30-150  (188) 

Pfister  and  others  1 975 

490-570 

150-250  (150) 

Whittaker  and  Niering  1975 

50-250  (350+) 

Weaver  and  Forcella  1 977 

Festuca  idahioensis 

235 

2.35 

Collins  and  Weaver  1978 

195 

1.95 

Weaver  and  Collins  1977 

147 

1.47 

Daubenmire  1970 

Festuca  scabrella 

152 

1.52 

Willms  and  others  1986 

Bouteioua  gracilis 

103 

1.03 

Weaver  1 983 

53-95 

0.5-1.0 

Hunt  and  others  1 988 

Atriplex  spp. 

28 

West  1983 

'Since  merchantable  production  and  standing  crop  are  emphasized  in  these  studies,  one  might  expect  the  figures  to  be  50  to  66  percent  of  those  reported  in 
studies  of  total  production  (Weaver  and  Forcella  1977). 

Where  production  and  standing  crop  are  expressed  volumetrically,  masses  were  calculated  using  specific  gravities  of  ABLA=0.38,  PIAL=0.40,  PIEL=0.35, 
PSME=0.45,  PIPO=0.43,  PICO=0.38,  THPL=0.33,  TSHE=0.41  (U.S.  FPL  1974). 


(starvation)  may  exclude,  from  an  altitudinal  zone,  plants 
of  lower  zones.  Average  growing  season  temperatures  in 
the  alpine  (and  for  most  zones  below  it)  are  distinctly 
lower  than  temperatures  in  the  vegetation  zone  below 
(fig.  3).  This  correlation  suggests  the  possibility  of — but 
does  not  prove — significant  effects  of  growing  season 
temperature.  In  support  of  this  h3T)othesis,  production- 
temperature  relations  (discussed  below)  suggest  that  the 
vegetation  of  each  altitudinal  zone  is  especially  adapted 
to  temperatures  occurring  in  its  zone.  In  this  regard,  it 
is  satisfying  to  see  that  optimum  temperatures  for  white- 
bark  pine  photosynthesis  (20-25  °C,  Jacobs  and  Weaver 
1990)  are  similar  to  summer  maximum  temperatures 
(10-25  °C,  Weaver  1990). 

Graphs  of  estimated  soil  water  availability  against  veg- 
etation type  (fig.  4)  indicate  little  or  no  stress  in  alpine, 
whitebark,  subalpine  fir,  and  Douglas-fir  zones.  On  this 
basis  one  might  exclude  droiight  as  a  factor  determining 
whitebark's  upslope  limit.  Due  to  two  modifying  factors, 


this  conclusion  seems  premature.  First,  as  one  moves 
upslope  from  forest  to  alpine  (or  moimtain  meadow), 
winds  increase  and,  with  increasing  wind,  drought  in- 
creases due  to  reduced  water  availability — snow  may  be 
blown  off  site  (Daubenmire  1981)  and  water  in  uninsu- 
lated soils  freezes — and  increasing  water  loss — through 
abrasion  of  cuticle  (Hadley  and  Smith  1987)  and  thinning 
of  the  leafs  boundary  layer  (Gates  and  Papian  1971;  Nobel 
1983).  I  see  winter  wind  effects  as  probable  controllers  be- 
cause trees  invading  (or  planted  into)  mountain  meadows 
are  more  often  desiccated  in  winter  than  stunmer.  The 
wind  effect  hypothesis  is  consistent  with  the  fact  that 
groups  of  trees  sometimes  invade  sites  that  individuals 
cannot  invade  alone  (Armand  1992;  Tranquillini  1979).  In 
such  situations  adjacent  trees  are  more  likely  sheltering 
each  other  from  drying  wind  than  from  lethal  low  tempera- 
tures. Second,  as  one  moves  upslope  the  average  soil  be- 
comes progressively  better  drained  (Weaver  1979),  and 
thus  effective  precipitation  is  a  smaller  fraction  of  total 


145 


10      20  30 


ECOSYSTEM  TYPE 


Figure  1 — Midwinter  low  temperatures  in  10  Rocky  Mountain  ecosystems.  Low  temperatures  (°C 
and  °F)  are  represented  by  absolute  winter  minimums  (A,  ABSMN)  and  average  January  minima 
(B,  AVJMN).  Ecosystem  types  range  altitudinally  from  alpine  (2),  down  through  forests  (8-20)  to 
grass  and  shrublands  (26-36);  specifically,  they  are  alpine  (2),  whitebark  pine  (8),  subalpine  fir  (10), 
Douglas-fir  (16),  cedar-hemlock  (18),  ponderosa  pine  (20),  Idaho  fescue  (26),  bluebunch  wheatgrass 
(28),  gramagrass  (30),  and  desert  shrub  (36).  Large  and  small  circles  represent  data  gathered  east 
and  west  of  the  Rockies,  respectively.  Lines  connect  median  values. 


30      40'         '0  10 

ECOSYSTEM  TYPE 


Figure  2— The  coldest  early  spring  (A,  SPR-FRST)  and  late  fall  (8,  FALL-FRST)  frosts  (°C 
and  "F)  in  10  Rocky  Mountain  ecosystems.  Ecosystems,  symbols,  and  lines  are  as  in  figure  1. 


O         10       20       30       40  O        10       20       30  40 

ECOSYSTEM  TYPE 

Figure  3 — Average  growing  season  temperatures  in  1 0  Rocky  Mountain  ecosystems.  The  first  graph 
(A,  TEMP  gs)  gives  the  average  growing  season  temperature  (°C  and  "F).  Q  gs  in  the  second  graph 
(B)  considers  the  exponential  response  of  physiological  processes  to  increasing  temperature  (Q,q). 
Ecosystem  types,  symbols,  and  lines  are  as  in  figure  1 . 


146 


precipitation  at  higher  than  lower  altitudes;  that  is,  the 
Walter  drought  indices  presented  in  figure  4  probably  over- 
estimate water  availability  on  (high-altitude)  thin-soil  sites. 

Downward  Limits 

I  speculate  that  downward  limits  of  whitebark  pine 
might  be  determined  by  heat,  drought,  or  competition. 
The  following  paragraphs  consider  these  hypotheses. 

As  one  moves  downslope  maximum  temperatures 
(fig.  5)  rise.  That  this  correlate  of  tree  disappearance  is 
not  controlling  is  suggested  by  three  facts.  Summer  aver- 
ages in  vegetation  zones  below  whitebark  pine  (10-15  °C, 


fig.  3)  are  well  below  the  whitebark  pine  optimimi 
(20-25  °C,  Jacobs  and  Weaver  1990),  so  downward  migra- 
tion might  actually  improve  production.  July  maxima 
(20-27  °C,  fig.  5)  in  zones  below  are  near  the  whitebark 
pine  optimvuu  temperature  (20-25  °C,  Jacobs  and  Weaver 
1990)  and  thus  should  not  be  damaging.  Long-term 
maxima  (30-42  °C,  fig.  5)  in  zones  below  do  not  eliminate 
net  photosynthesis  (Jacobs  and  Weaver  1990)  and  thus 
are  probably  not  lethal.  In  addition,  if  whitebark  pine 
trees  are  well  watered,  they  grow  well  as  lawn  trees  in  the 
Agropyron  spicatum  zone  (for  example,  Belgrade,  MT) 
where  temperatures  are  far  higher  than  those  fovmd  on 
sites  whitebark  naturally  occupies. 


ECOSYSTEM  TYPE 

Figure  4 — Drought  severity  in  10  Rocky  Mountain  ecosystems.  The  first  graph  (A) 
gives  the  number  of  dry  months  (DRT-MNTHS).  The  second  (B)  indexes  the  an- 
nual water  deficiency  (W-DFCT,  mm).  Both  were  calculated  after  Walter  (1973). 
The  ecosystem  types,  symbols,  and  lines  are  as  in  figure  1 . 


0       10     20     30     40  0       10     20     30  40 

ECOSYSTEM  TYPE 

Figure  5 — High  temperatures  (°C  and  °F)  recorded  in  10  Rocky  Mountain  ecosys- 
tems. The  first  graph  (A)  represents  absolute  summer  highs  (ABSMX).  The  sec- 
ond (B)  gives  average  July  maxima  (AVJMX).  The  ecosystem  types,  symbols,  and 
lines  are  as  in  figure  1 . 


147 


ECOSYSTEM  TYPE 


Figure  6— Biomass  production  in  10  Rocky  Mountain  ecosystems  (see  also  table  2).  The 
first  graph  (A)  presents  harvestable  production  (HPRD,  g/m^/yr).  The  second  (B)  gives  to- 
tal production  (TPRD,  g/m^/yr).  The  ecosystem  types,  symbols,  and  lines  are  as  in 
figure  1 . 


Drought  likely  sets  the  lower  physiological  limit 
(Hutchinson  1958)  for  whitebark  pine's  downward  exten- 
sion. Walter's  (1973)  index  suggests  the  drought  months 
and  water  deficits  are  near  zero  in  high  forests  (whitebark 
pine,  subalpine  fir,  and  Douglas-fir),  greater  in  ponderosa 
pine  forests  and  grasslands,  and  still  greater  in  Atriplex 
shrublands  (fig.  4).  If  (as  noted  above)  wind  decreases 
and  waterholding  capacity  of  soils  increases  downslope, 
water  conditions  for  whitebark  production  improve  as  one 
moves  downslope;  this  is  consistent  with  the  fact  that 
whitebarks  growing  in  the  subalpine  fir  zone  can  be  stately 
timber  producers  (Pfister  and  others  1977;  Weaver  and 
Dale  1977).  Water  deficits  remain  slight  down  through  the 
subalpine  fir  and  possibly  the  Douglas-fir  zones.  The  com- 
plete absence  of  whitebarks  in  the  ponderosa  pine  and 
grassland  zones,  despite  possible  nutcracker  dispersal 
(Lanner  1990;  Tomback  and  others  1990),  supports  my 
doubt  that  the  tree  is  capable  of  growing  in  the  ponderosa 
pine  and  drier  grassland-desert  zones.  Similarly,  in  the 
geographic  dimension,  the  southern  and  eastern  limits  of 
whitebark  pine  seem  to  appear  where  precipitation  in  its 
altitudinal  zone  becomes  too  low  (Arno  and  Weaver  1990). 

The  lower  realized  limit  (Hutchinson  1958)  of  white- 
bark is  set  by  competition  with  subalpine  fir  and  lodge- 
pole  pine  rather  than  drought.  In  our  region,  evidence  for 
this  fact  lies  in  the  tree's  existence  and  good  growth  on 
sites  in  the  subalpine  fir  zone  (but  apparently  not  in  the 
Douglas-fir  zone)  where  subalpine  fir  and  lodgepole  com- 
petition have  been  removed  by  clearcutting  or  fire.  The 
lower  limit  of  whitebark  is,  then,  set  by  those  factors — 
likely  drought  (thin  soils  and  wind)  and  low  summer  tem- 
perature— that  exclude  subalpine  fir  and  lodgepole  pine 
from  higher  sites.  In  the  geographic  dimension,  to  the 
north  and  west  where  rainfall  is  higher,  whitebark  is  also 
unable  to  form  pure  stands  or  disappears  (Arno  and 
Weaver  1990).  (European  foresters  ask  why  we  don't 
eliminate  competitive  "natives"  [subalpine  fir  and  lodgepole 
pine]  from  their  zone  to  allow  expansion  of  whitebark  pine 
and  increased  production  of  quality  white  pine  lumber.) 


Production 

As  one  moves  from  dry  grasslands  to  moist  forests  and 
upward  to  alpine  sites  one  sees  production  rise  and  fall 
again.  Harvestable  production — merchantable  standing 
crop  divided  by  years  in  its  production — rises  from  dry 
grasslands  (100  g/mVyr)  to  moist  forests  (300  g/mVyr)  and 
falls  through  high  forests  to  the  alpine  (100  g/mVyr,  fig.  6, 
table  2).  Available  data  suggest  that  total  aboveground 
production  (total  increment  ^ro?lyr)  also  increases  fi-om 
dry  grass  and  shrublands  (50-100  g/mVjr)  to  warm  forests 
(over  500  g/mVyr)  and  falls  to  the  alpine  (100  g/mV5T, 
fig.  6,  table  1).  The  trend  is  parallel,  but — at  least  in  for- 
ests— total  production  is  far  higher  than  harvestable  pro- 
duction. Three  factors  may  contribute  to  this  difference: 
(1)  Much  of  the  production,  over  the  life  of  a  stand,  is  lost 
to  needle  drop,  self  pruning,  and  natural  thinning;  this 
material  decomposes  and  is  recycled.  The  total  produc- 
tion rates  reported  are  deceptively  high  because  they  in- 
clude nutrients  being  recycled  to  the  atmosphere  and  soil; 
that  is,  this  production  consists  of  an  actual  production 
component  based  on  nutrient  import  (weathering  and  at- 
mospheric delivery)  and  a  restructuring  component  based 
on  recycling  of  nutrients  from  "obsolete"  leaves,  branches, 
and  trees.  (2)  Only  33  to  50  percent  of  the  material  extant 
at  harvest  is  merchantable  (Weaver  and  Forcella  1977). 
(3)  Most  of  the  total  production  data  comes  from  areas 
outside  our  region.  If  the  climates  of  Douglas-fir,  cedar- 
hemlock,  and  ponderosa  pine  stands  from  which  produc- 
tion data  are  reported  are  moister  or  warmer  than  the 
Montana  stands  in  which  the  harvestable  productivity 
data  were  gathered,  somewhat  higher  productivities 
might  be  expected. 

I  expect  productivity  to  be  controlled,  like  vegetation 
distribution,  by  climate.  In  the  following  paragraphs  I 
will  explore  the  h3rpothesis  that  production  depends  on 
growing  season  length  and  the  warmth  of  that  season. 

Growing  season  length  might  be  indexed  as  the  number 
of  months  where  daily  average  temperatures  are  above 


148 


Table  3 —  Relationship  of  production  (g/m^/yr)  to  climatic  factors 


Harvestable  production' 


HP 

66.00 

+ 

12.8000 

WS 

p  =  0.50 

=  0.05 

-171.00 

+ 

70.0000 

GS 

p  =  0.00 

=  0.86 

34.00 

+ 

1.7700 

GSxT 

p  =  0.04 

/■2  =  0.44 

-38.00 

+ 

14.4000 

GSxQ 

p  =  0.01 

=  0.58 

Log  HP  = 

0.78 

wo 

n  -  n  nn 

f2  _  n  7fi 

1  75 

n  nnfin 

n  —  n  1 5 

f2  —  n  97 

1.44 

+ 

0.0530 

GSxQ 

p  =  0.03 

r2  =  0.44 

tal  Droduction^ 

TP 

-880.00 

+ 

321.0000 

GS 

p  =  0.16 

r2  =  0.22 

50.00 

+ 

8.3200 

GSxT 

p  =  0.32 

=  0.12 

-398.00 

+ 

77.0000 

GSxQ 

p  =  0.19 

r2  =  0.20 

Log  TP  = 

0.40 

+ 

0.4700 

GS 

p  =  0.03 

=  0.45 

1.96 

+ 

0.0086 

GSxT 

p  =  0.32 

r2  =  0.12 

1.47 

+ 

0.0820 

GSxQ 

p  =  0.22 

^  =  0.22 

'Harvestable  production  (HP,  lop  diameter  greater  than  10  cm)  and  the  log  of  harvestable  production 
(log  HP)  are  regressed  against  warm  season  (WS),  warm-moist  growing  season  (GS),  the  product  of 
warm-moist  growing  season  and  average  growing  season  temperature  (GSxT),  and  the  product  of 
wann-moist  growing  season  and  Q,o-based  growth  support  units  (GSxQ). 

n"otal  production  (TP)  and  log  total  production  (log  TP)  are  also  regressed  against  GS,  GSxT,  and 
GSxQ. 


0  °C.  The  rationale  is  that,  iinless  air  temperatiires  are 
above  0  °C,  exposed  soils  will  be  frozen  (Weaver,  in  prepa- 
ration) and  frozen  soils  will  deliver  little  water  and  nutri- 
ents. Because  a  regression  of  production  against  this  in- 
dex of  growing  season  is  so  loose  (r^  =  0.05,  table  3,  fig.  7), 
the  hypothesis  that  warm-season  length  controls  produc- 
tion is  inadequate. 

The  previous  index  may  fail  because,  at  least  at  lower 
altitudes,  growing  season  is  limited  in  late  summer  by  a 
lack  of  soil  water,  rather  than  temperatiire.  If  so,  the 
growing  season  index  would  be  improved  as  an  index  of 
production  by  subtracting  soil-dry  months  from  total  soil- 
warm  months.  As  above,  I  index  drought  as  periods  when 
precipitation  (mmy2)  is  less  than  average  temperature  (C, 
after  Walter  1973).  Regressions  of  production  against 
months  without  either  temperature  or  moisture  stress  ex- 
plain most  of  the  variance  in  harvestable  production  (r^  = 
0.86,  table  3,  fig.  7).  Season  length  is  a  weaker  correlate 
of  total  production  (r^  =  0.22).  I  offer  two  unsatifying 
hypotheses  for  this  important  difference:  (1)  Growing  sea- 
son predicts  harvestable  production  better  than  total  pro- 
duction because  photosynthates  are  allocated  first  to 
canopy  maintainence  (so  this  component  does  not  vary 
much  among  types)  and  second  to  supporting  structure;  if 
so,  the  relationship  between  season  length  and  tnmk  pro- 
duction is  clarified  by  omitting  the  common  leaf-twig  pro- 
duction. (2)  If  total  production  were  measured  in  slightly 
more  favorable  segments  of  the  environmental  types  than 
harvestable  production,  the  iinderstatement  of  season 
length  would  weaken  the  relationship. 

If  all  vegetation — grassland  to  forest  to  alpine — had  a 
single  temperature  response  curve,  we  would  expect  pro- 
duction in  "warm  moist  soil  days"  to  rise  exponentially 
with  increasing  temperatiu-e.  This  expectation  is  based 


on  the  rise  in  chemical  reaction  rates  with  rising  tem- 
perature, on  the  resultant  physiological  response  curves 
(Larcher  1975),  on  degree-day  prediction  of  plant  phenol- 
ogy (Chang  1960),  and  on  the  prediction  that  degree-day 
formulae  might  be  improved  by  use  of  temperatiire  aver- 
ages computed  from  a  nonlinear  Q^^  relationship  (Larcher 
1975).  Thus,  we  expect  production  to  be  better  correlated 
with  the  product  of  growing  season  days  x  temperature 
than  with  growing  season  days  alone.  In  fact,  whether 
we  use  our  average  growing  season  temperature  or  grow- 
ing season  Q  index,  the  relationship  is  poorer  for  both 
harvestable  (r^=  0.44)  and  total  (r^=  0.12)  production. 
Since  the  basic  physiological  response  is  a  physiochemical 
necessity,  I  deduce  first  that,  while  the  vegetation  of  each 
environmental  zone  does  respond  to  temperature,  it  is 
specifically  adapted  to  temperattires  in  that  zone  and,  sec- 
ond, that  temperature  adaptation  distinct  to  each  zone 
eliminates  temperature  fi'om  the  predictive  equation.  I 
doubt  that  temperature  would  be  eliminated  if  adaptation 
were  eliminated  by  using  a  genetically  homogeneous  veg- 
etation type;  for  example,  if  the  region  were  vegetated 
with  a  single  crop,  production  would  be  best  predicted  by 
the  product  of  growing  season  length  and  temperature. 
Since  competitiveness  of  a  tree  undoubtedly  depends  on 
its  productivity  relative  to  competitors,  this  conclusion 
supports  earlier  speculation  that  growing  season  tempera- 
ture adaptation  provides  at  least  one  basis  for  the 
existance  of  different  vegetation  zones. 

While  I  have  argued  that  production  in  a  zone  is  best 
predicted  from  the  length  of  its  warm  moist  growing  sea- 
son, I  expect  the  standing  crop  of  mature  accvmaulative 
(woody)  vegetation  (table  2)  to  be  largely  determined  by 
site  fertility.  Accmnulation — slow  or  fast — depends  on 
photosynthesis  minus  respiration.  Synthesis  depends  on 


149 


light,  warmth,  water  (open  stomates),  and  nutrients.  In 
the  absence  of  stand  destruction,  annual  inputs  of  light, 
warmth,  water,  and  elements  with  atmospheric  cycles 
(for  example,  C,  H,  O,  N)  should  support  eternal  accumu- 
lation. Limited  supplies  of  nutrients  with  geologic  cycles 
(for  example,  P,  K,  Ca)  will  halt  accumulation  (Weaver 
1978).  It  is  also  argued  (Odum  1969)  that  accvunulation 
is  halted  when  respiration  equals  production.  Thus,  in 
young  trees  net  production  is  high  because  the  ratio  of 
photosynthetic  mass  to  respiratory  mass  is  high,  in  longer 
stemmed  trees  efflcency  declines  as  the  respiratory  load 
increases,  and  growth  ceases  when  respiration  equals 
gross  production;  heartwood's  contribution  to  this  rela- 
tionship is  in  proportion  to  its  inertness.  While  the  respi- 
ratory factor  undoubtedly  contributes  to  observed  declines 
in  production  with  stand  height  (and  age)  (Weaver  and 
others  1990),  I  see  nutrient  supplies  as  more  limiting  in 
high  forests  because  the  trees  are  far  shorter  (and  there- 
fore more  efficient)  than  productive  trees  of  lower  alti- 
tudes. In  contrast,  standing  crops  in  less  accumulative 
(herbaceous)  vegetation  are  determined  by  production  oc- 
curring in  one  growing  season,  and  while  it  could  be  de- 
termined by  supplies  of  a  geologically  cycling  nutrient, 
standing  crop  is  more  likely  determined  by  a  bulk  re- 
source like  light  (unlikely),  warmth,  water,  or  an  atmo- 
spheric nutrient  like  nitrogen. 


CONCLUSIONS 

Whitebark  pine  vegetation  is  likely  excluded  from  the  al- 
pine zone  by  cool  growing  season  temperatiires  or  droughts 
occurring  most  likely  in  winter.  It  seems  unlikely  that 
freezes  of  late  fall,  midwinter,  or  early  spring  exclude  the 
tree  from  higher  sites. 

The  lower  physiological  limit  of  whitebark  pine  is  likely 
set  by  drought.  Its  lower  realized  limit  is  likely  due  to  com- 
petition with  lodgepole  pine  or  subalpine  fir.  By  control  of 
competition,  managers  could  probably  extend  whitebark's 
range  downslope. 

The  distribution  of  other  dominant  plants  on  the  tem- 
perature water  gradient  may  be  similarly  controlled. 

Annual  production  is  strongly  related  (H  =  0.86)  to  the 
number  of  growing  season  days.  The  lack  of  correlation 
with  temperature  suggests  that  plants  of  any  vegetation 
zone  are  adapted  to  temperature  conditions  peculiar  to 
that  zone.  Mature  standing  crops  of  woody  vegetation  are 
more  likely  determined  by  nonatmospheric  nutrients  than 
temperature  and  precipitation.  Potential  standing  crops 
of  herbaceous  vegetation,  on  the  other  hand,  are  more 
likely  limited  by  bulk  resources  such  as  warmth,  water, 
carbon,  or  nitrogen. 


3  6 

WM  SEASON,  mo 


20  40 

GS  X  T  (moxC  5) 


\ 

o 

cr 

CL 
X 


r2=  0.85 


B 


J  L 


J  L 


J  L 


3  6 

GRO  SEASON,  mo 


GS  X  Q  (moxQ) 


Figure  7 — The  relationship  of  harvestable  production  (HPROD,  g/m^/yr)  to  season 
length  and  temperature:  (A)  production  vs.  warm  season  (WM-SEASON),  (B)  produc- 
tion vs.  warm-moist  season  (GRO-SEASON),  (C)  production  vs.  warm-moist  season 
(GS)  X  average  summer  temperature  above  5  °C  (C-5),  (D)  production  vs.  warm  moist 
season  (GS)  x  growth  support  units  (Qg^).  The  ecosystem  types  are  as  in  figure  1 . 


150 


ACKNOWLEDGMENTS 

I  thank  colleagues  who  helped  me  locate  weather  sta- 
tions representative  of  major  environmental  types  of  the  re- 
gion (S.  Amo,  D.  Despain,  J.  Habeck,  R.  Rankin,  and  others 
acknowledged  in  Weaver  1980  and  1990),  acquire  the  data 
(J.  Caprio),  and  choose  parameters  for  study  (J.  Pickett  and 
J.  Brown).  The  critical  review  of  P.  Burgeron,  D.  Despain, 
S.  Harvey,  W.  Moir,  and  D.  Perry  is  greatly  appreciated.  I 
am  also  grateful  for  the  support  provided  by  Intermoimtain 
Research  Station,  USDA  Forest  Service  (INT-92720- 
RJVA). 

REFERENCES 

Aplet,  G.;  Smith,  F.;  Laven,  R.  1989.  Stemwood  biomass 
and  production  during  spruce-fir  stand  development. 
Journal  of  Ecology.  77:  70-77. 

Armand,  A.  1992.  Sharp  and  gradual  motmtain  timber- 
lines  as  a  result  of  species  interaction.  In:  Hansen,  A.; 
DiCastri,  F.  1992.  Landscape  boundaries.  New  York: 
Springer  Verlag:  360-378. 

Amo,  S.  1984.  Timber  line — moimtain  and  arctic  forest 
frontiers.  Seattle:  Moimtaineers.  304  p. 

Amo,  S.;  Weaver,  T.  1990.  Whitebark  pine  commimity 
types  and  their  pattem  in  the  landscape.  In:  Schmidt, 
W.;  McDonald,  K,  comps.  Proceedings —  symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-moimtain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermovmtain  Research  Station:  97-105. 

Burke,  M.;  Gusta,  L.;  Quamme,  H.;  Weiser,  C.;  Li,  P. 
1976.  Freezing  injxiry  in  plants.  Annual  Review  of  Plant 
Physiology.  27:  507-528. 

Chang,  J.  1968.  Climate  and  agriculture.  Chicago:  Aldine. 
304  p. 

Collins,  D.;  Weaver,  T.  1978.  Effects  of  summer  weather 
modification  (irrigation  in  Festuca  idahoensis-Agro- 
pyron  spicatum  grasslands.  Jovu-nal  of  Range  Manage- 
ment. 31:  264-269. 

Daubenmire,  R.  1943.  Vegetation  zonation  in  the  Rocky 
Mountains.  Botanical  Review.  9:  325-393. 

Daubenmire,  R.  1956.  Climate  as  a  determinant  of  vegeta- 
tion distibution  in  eastern  Washington  and  northern 
Idaho.  Ecological  Monographs.  26:  131-154. 

Daubenmire,  R.  1970.  Steppe  vegetation  of  Washington. 
Bull.  62.  Pullman,  WA:  Washington  Agricultural  Ex- 
periment Station.  131  p. 

Daubenmire,  R.  1981.  Subalpine  parks  associated  with 
snow  transfer  in  the  mountains  of  northern  Idaho  and 
eastem  Washington.  Northwest  Science.  55:  124-135. 

Daubenmire,  R.;  Daubenmire,  J.  1968.  Forest  vegetation 
of  eastern  Washington  and  northern  Idaho.  Bull.  60. 
Pullman,  WA:  Washington  Agricultural  Experiment 
Station.  104  p. 

Forcella,  F.;  Weaver,  T.  1977.  Biomass  and  productivity  of 
the  subalpine  Pinus  albicaulis/Vaccinium  scoparium  as- 
sociation, MT  USA.  Vegetatio.  35:  95-105. 

Gams,  H.  1931.  Die  klimatische  Begrenzimg  von  Pflanzen 
arealen  und  die  Verteilvmg  der  hygrischen  Konti- 
nentalitat  in  den  Alpen.  Zietschrit  der  Gesellschafl  fur 
Erdkunde.  9:  32-68. 


Gates,  D.;  Papian,  L.  1971.  Atlas  of  energy  budgets  of 
plant  leaves.  London  and  New  York:  Academic  Press. 
277  p. 

Geiger,  R.  1965.  The  climate  near  the  ground.  Cambridge, 

MA:  Harvard  University  Press.  611  p. 
Hadley,  J.;  Smith,  W.  1987.  Influence  of  knunmholz 

microclimate  on  needle  physiology  and  survival. 

Oecologia.  73:  82-90. 
Hanley,  D.  1976.  Tree  biomass  and  production  estimates 

for  three  habitat  types  in  northern  Idaho.  Bull.  14. 

Moscow,  ID:  University  of  Idaho,  College  of  Forestry. 

16  p. 

Hunt,  W.;  Ingham,  E.;  Coleman,  D.;  ElHott,  E.;  Reid,  P. 
1988.  Nitrogen  limitation  of  production  and  decomposi- 
tion in  prairie,  moimtain  meadow,  and  pine  forest. 
Ecology.  69:  1009-1016. 

Holdridge,  L.  1967.  Determination  of  world  plant  forma- 
tions from  simple  climatic  data.  Science.  130:  572. 

Huschle,  G.;  Hironaka,  M.  1980.  Classification  and  ordi- 
nation of  serai  plant  communities.  Journal  of  Range 
Management.  33:  179-182. 

Hutchinson,  G.  1958.  Concluding  remarks.  Cold  Springs 
Harbor  Symposium  in  Quantiative  Biology.  22:  415-427. 

Jacobs,  J.;  Weaver,  T.  1990.  Effect  of  temperature  and 
temperature  preconditioning  on  seedlimg  performance 
of  whitebark  pine.  In:  Schmidt,  W.;  McDonald,  K, 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S. 
Department  of  Agriculture,  Forest  Service,  Intermoun- 
tain  Research  Station:  134-139. 

Kuchler,  A.  1964.  Potential  vegetation  of  the  contermi- 
nous US.  Spec.  Publ.  36.  New  York:  American  Geo- 
graphical Society.  155  p. 

Landis,  T.;  Mogren,  E.  1975.  Tree  strata  biomass  of  sub- 
alpine spmce-fir  stands  in  SW  Colorado.  Forest  Science. 
21:  9-12. 

Lanner,  R.  1980.  Biology,  taxonomy,  evolution,  and  geog- 
raphy of  the  stone  pines  of  the  world.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  14-24. 

Larcher,  W.  1975.  Physiological  plant  ecology.  New  York: 
Springer.  252  p. 

Neilson,  R.  1986.  High  resolution  climatic  analysis  and 
southwest  biogeography.  Science.  232:  27-33. 

Nielson,  R.  1992.  Toward  a  rule  based  biome  model.  Land- 
scape Ecology.  7:  27-43. 

Nobel,  P.  1983.  Biophysical  plant  physiology.  San 
Francisco:  Freeman.  608  p. 

Penman,  H.  1949.  The  dependence  of  transpiration  on 
weather  and  soil  conditions.  Journal  of  Soil  Science. 
1:  74-89. 

Pfister,  R.;  Kovalchik,  B.;  Amo,  S.;  Presby,  R.  1977.  For- 
est habitat  types  of  Montana.  Gen.  Tech.  Rep.  INT-34. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Forest  and  Range  Experiment  Sta- 
tion. 174  p. 


151 


Rodin,  L.;  Bazilevich,  N.  1967.  Production  and  mineral  cy- 
cling in  terrestrial  vegetation.  London:  Oliver  and  Boyd. 
288  p. 

Salisbury,  F.;  Ross,  C.  1992.  Plant  physiology.  Belmont, 
CA:  Wadsworth.  682  p. 

Scott,  D.;  Billings,  W.  1964.  Effects  of  environmental  fac- 
tors on  standing  crop  and  production  of  an  alpine  tun- 
dra. Ecological  Monographs.  34:  243-270. 

Stephenson,  N.  1990.  Climatic  control  of  vegetation  distri- 
bution: the  role  of  water  balance.  American  Naturalist. 
135:  649-670. 

Thilenius,  J.;  Smith,  D.;  Brown,  G.  1974.  Effect  of  2,4-D 
on  composition  and  production  of  an  alpine  plant  com- 
munity in  Wyoming.  Journal  of  Range  Management. 
27:  140-142. 

Thornthwaite,  C.  1948.  An  approach  toward  a  rational 
classification  of  climate.  Geographical  Review.  38: 
55-94. 

Tomback,  D.;  Hoffmann,  L.;  Sund,  S.  1990.  Coevolution  of 
whitebark  pine  and  nutcrackers:  implications  for  forest 
regeneration.  In:  Schmidt,  W.;  McDonald,  K.,  comps. 
Proceedings — sjrmposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source. Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  De- 
partment of  Agriculture,  Forest  Service,  Intermountain 
Research  Station:  118-129. 

Tranquillini,  W.  1979.  Physiological  ecology  of  the  alpine 
timberline.  Tree  existence  at  high  altitude  with  special 
reference  to  the  European  Alps.  New  York:  Springer. 
137  p. 

U.S.  FPL.  1974.  Wood  handbook:  wood  as  an  engineering 
material.  Agric.  Handb.  72.  Washington,  DC:  U.S.  De- 
partment of  Agriculture,  Forest  Service,  Forest  Prod- 
ucts Laboratory,  [n.p.]. 

USGS.  1970.  The  national  atlas.  Washington,  DC:  U.S. 
Department  of  the  Interior,  Geological  Survey.  417  p. 

Walter,  H.  1973.  Vegetation  of  the  earth  in  relation  to  cli- 
mate and  ecophysiological  conditions.  New  York: 
Springer.  237  p. 

Weaver,  T.  1978.  Changes  in  soils  along  a  vegetation- 
altitudinal  gradient  of  the  Northern  Rocky  Mountains. 
In:  Youngberg,  C,  ed.  Forest  soils  and  land  use.  Pro- 
ceedings; 5th  North  American  Soils  Conference.  Fort 
Collins,  CO:  Colorado  State  University,  Forestry 
Department:  14-29. 


Weaver,  T.  1980.  Climates  of  vegetation  types  of  the 
northern  Rocky  Mountains  and  adjacent  plains.  Ameri- 
can Midland  Naturalist.  103:  392-398. 

Weaver,  T.  1983.  Yield  response  of  high  plains  grasslands 
to  water  enrichment,  three  phases.  In:  Holman,  L.; 
Knutsen,  G.  State  of  Montana  activities  in  the  High 
Plains  Cooperative  Program  1981-1983.  Helena,  MT: 
Montana  Department  of  Natural  Resources,  Water 
Resources  Division:  77-94. 

Weaver,  T.  1990.  Climates  of  subalpine  pine  woodlands. 
In:  Schmidt,  W.;  McDonald,  K.,  comps.  Proceedings — 
S3mtiposium  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
72-79. 

Weaver,  T.;  Collins,  D.  1977.  Effects  of  weather  modifica- 
tion (increased  snow  pack)  on  Festuca  idahoensis  mead- 
ows. Journal  of  Range  Management.  30:  451-456. 

Weaver,  T.;  Dale,  D.  1974.  Pinus  albicaulis  in  central 
Montana:  environment,  vegetation,  and  production. 
American  Midland  Naturalist.  92:  222-229. 

Weaver,  T.;  Forcella,  F.  1977.  Biomass  of  fifty  conifer  for- 
ests and  nutrient  exports  associated  with  their  harvest. 
Great  Basin  Naturahst.  37:  395-401. 

Weaver,  T.;  Forcella,  F.;  Dale,  D.  1990.  Stand  develop- 
ment in  whitebark  pine  woodlands.  In:  Schmidt,  W.; 
McDonald,  K.,  comps.  Proceedings — symposium  on 
whitebark  pine  ecosystems:  ecology  and  management 
of  a  high-mountain  resource.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  151-155. 

West,  N.  1983.  Temperate  deserts  and  semideserts:  Inter- 
mountain salt  desert  shrubland.  Amsterdam:  Elsevier: 
382. 

Whittaker,  R.  1975.  Communities  and  ecosystems.  New 

York:  Macmillan.  385  p. 
Whittaker,  R.;  Niering,  W.  1975.  Vegetation  of  the  Santa 

Catalina  Mountains,  Arizona.  V.  Biomass,  production 

and  diversity  along  the  elevation  gradient.  Ecology.  56: 

771-790. 

Willms,  W.;  Smoliak,  S.;  Bailey,  A.  1986.  Herbage  produc- 
tion following  litter  removal  on  Alberta  native  grass- 
lands. Journal  of  Range  Management.  39:  536-540. 


152 


Regeneration 


153 


A 


VARIATION  IN  SIZE  AND  WEIGHT  OF 
CONES  AND  SEEDS  IN  FOUR  NATURAL 
POPULATIONS  OF  CARPATHIAN 
STONE  PINE 

I.  Blada 
N.  Popescu 


Abstract — High  variation  was  found  within  populations  of 
stone  pine  (Pinus  cembra  L.)  for  seeds  per  cone  and  seeds  per 
cone  weight,  but  middle  high  for  1,000-seed  weight  and  low  for 
cone  length  and  cone  diameter.  The  cone  length,  cone  diameter, 
seeds  per  cone,  seeds  per  cone  weight,  and  1,000-seed  weight 
means  were  4.7  cm,  4.2  cm,  37.8  cm,  9.2  g,  and  250  g,  respec- 
tively. Southern  populations  were  similar  in  cone  length  and 
cone  diameter,  but  not  in  seeds  per  cone,  seeds  per  cone  weight, 
and  l.OOO-seed  weight.  Cone  length  and  cone  diameter  from 
open-pollinated,  cross-pollinated  and  self-pollinated  cones  were 
similar,  while  seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight,  were  not.  All  traits  displayed  continuous  variation. 
Significant  correlations  were  found  among  all  traits,  except 
1,000-seed  weight. 


Stone  pine  (Pinus  cembra  L.)  is  naturally  distributed 
in  the  highest  forest  zone  of  the  Alps  and  Carpathian 
Mountains  (Critchfield  and  Little  1966;  Holzer  1975). 
In  the  Alps,  the  low-elevation  stands  range  between  1,100 
to  1,500  m,  but  the  main  zone  extends  between  1,700  to 
2,000  m  (Contini  and  Lavarelo  1982;  Holzer  1975),  while 
the  high-elevation  form  of  the  species  climbs  as  single 
trees  up  to  2,700  m  above  sea  level  (Moser  1960).  In 
Romania,  stone  pine  ranges  from  1,350  to  1,880  m  in  the 
northern  Carpathians  (Gubesch  1971)  and  from  1,350  to 
1,980  m  in  the  southern  Carpathians  (Beldie  1941;  Oarcea 
1966;  Tataranu  and  Costea  1952). 

Because  of  its  tolerance  for  low  temperatvires,  the  spe- 
cies is  very  important  for  reforestation  of  the  subalpine 
zone;  in  this  zone  it  is  also  important  on  watersheds,  for 
stabilizing  avalanche  areas,  and  for  reducing  the  effects 
of  flash  floods  (Holzer  1972). 

Stone  pine  has  a  particular  importance  for  the  silvicul- 
ture of  the  subalpine  zone  of  the  Carpathians.  For  this 
reason  a  genetic  improvement  program  with  both  intra- 
and  interspecific  crosses  is  being  carried  out  (Blada  1982, 
1990a);  some  results  have  been  published  (Blada  1987, 
1990b,  1992a)  or  are  in  preparation  (Blada  1992b). 

This  paper  reports  on  the  phenotypic  variation  in  size 
and  weight  of  cones  and  seeds  in  four  natural  populations 
of  Carpathian  stone  pine,  as  part  of  the  program. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

I.  Blada  and  N.  Popescu  are  Forest  Geneticists,  Forest  Research  Insti- 
tute, Bucharest  11,  Romania. 


MATERIALS  AND  METHODS 

Twenty  trees  were  sampled  at  random  in  each  of  the 
four  populations  listed  in  table  1.  In  mid-July  1991,  25 
cones  from  open  pollination  (OP)  on  each  of  the  20  trees 
(total  80  trees)  were  protected  against  the  mountain  jay 
(Nucifraga  caryocatactes  L.)  by  using  metal  net  bags  25 
by  20  cm  in  size. 

For  comparative  reasons,  cones  and  seeds  obtained  from 
controlled  cross-  and  self-pollinations  (CP,  SP)  in  the 
Gemenele  population  were  measured;  these  pollinations 
were  performed  in  July  1989  using  10  x  10  full  diallel 
mating  design  (Blada  1992b),  according  to  Griffing's 
(1956)  Experimental  Method  1. 

Cone  and  seed  measurements  were  taken  soon  afler 
their  collection  in  October  1990  for  controlled  cross- 
pollinated  and  self-pollinated  and  in  October  1991  for 
open-pollinated  cones.  Twenty-five  cones  were  measured 
from  each  tree  resulting  fi*om  open-pollinated  cones,  and 
20  cones  fi-om  each  combination  resulting  fi*om  controlled 
cross-pollinated  and  self-pollinated  cones.  Five  traits 
were  measured  as  shown  in  table  2. 

Using  data  from  measurements,  the  population  mean 
(x),  standard  deviation  (g),  mean  square  {&),  variation  co- 
efficient (VC),  range  of  variation  (Q),  and  correlation  coef- 
ficient (r)  were  calculated. 

WITHES-POPULATION  VARIATION 

Statistical  parameters  of  the  cones  and  seeds  of  the  four 
Carpathian  stone  pine  populations  listed  in  table  1  are 
summarized  in  table  3  and  figure  1. 


Table  1 — Geographic  distribution  of  Pinus  cembra  studied 
populations  in  the  Carpathians 


Population 

Latitude  N.  Longitude  E. 

Altitude 

Carpathian 
Range 

 Degrees  

Meters 

Gemenele 

45°35"  22°50' 

1,720 

Southern 

Stana  de  Rau 

45°25'  23°03' 

1,450 

Southern 

Pietrele 

45°23'  22°52' 

1,550 

Southern 

Lala 

47°33'  25°05' 

1,520 

Northern 

154 


Table  2 — Measured  traits 


Trait  Units  Symbols 


Cone  length  Centimeters  CL 

Cone  diameter  Centimeters  CD 

Seeds  per  cone  Number  SC 

Seeds  per  cone  weight  Grams  SCW 

1 ,000-seed  weight  Grams  1 ,000-SW 


The  following  main  results  were  obtained  for  the 
Gemenele,  Stana  de  Rau,  Pietrele,  and  Lala  populations 
(in  sequence): 

•  The  cone  length  mean  was  4.9  ±  0.4  cm,  4.7  ±  0.4  cm, 
4.9  ±  0.6  cm,  and  4.1  ±  0.4  cm,  respectively. 

•  The  cone  diameter  mean  was  4.2  ±  0.3  cm,  4.1  ±  0.3  cm, 
4.1  ±  0.3  cm,  and  3.5  ±  0.2  cm,  respectively. 

•  The  seeds  per  cone  mean  was  40.9  ±  9.6,  36.7  ±  8.1, 
52.6  ±  13.3,  and  21.1  ±  7.1,  respectively. 

•  The  seeds  per  cone  weight  was  9.8  ±  2.7  g,  9.0  ±  2.5  g, 
12.5  ±  3.9  g,  and  5.6  ±  1.7  g,  respectively. 


Table  3 — Mean  (x),  phenotypic  standard  deviation  (a),  mean  square 
(cf),  variation  coefficient  {VQ,  range  of  variation  (O) 


Parameters  (open  pollination) 


Trait 

x±  a 

VC 

Q 

Gemenele  population 

CL 

4.9 

±  0.4 

0.15 

7.9 

4.1 

5.8 

CD 

4.2 

+  0.3 

0.11 

7.9 

3.6 

5.0 

SC 

40.9 

±  9.6 

91.31 

23.3 

22.5 

60.2 

SCW 

9.8 

±  2.7 

7.50 

28.0 

4.6 

18.1 

1,000-SW 

238.0 

±  47.4 

2,243 

19.9 

170 

352 

Stana  de  Rau  population 

CL 

4.7 

±  0.4 

0.18 

9.1 

3.8 

5.6 

CD 

4.1 

±  0.3 

0.08 

7.0 

3.7 

4.9 

SC 

36.7 

±  8.1 

65.67 

22.1 

21.7 

55.4 

SCW 

9.0 

±  2.5 

6.05 

27.3 

5.5 

14.1 

1,000-SW 

252 

±  37.6 

1,412 

14.9 

175 

293 

Pietrele  population 

CL 

4.9 

±  0.6 

0.34 

11.7 

4.1 

6.0 

CD 

4.1 

±  0.3 

0.10 

7.7 

3.4 

4.7 

SC 

52.6 

±  13.3 

177.31 

25.3 

31.2 

90.4 

SCW 

12.5 

±  3.9 

15.27 

31.3 

6.2 

23.1 

1 ,000-SW 

238 

±  60.9 

3,712 

25.6 

138 

428 

Lala  population 

CL 

4.1 

±  0.4 

0.17 

9.9 

3.7 

5.1 

CD 

3.5 

±  0.2 

0.04 

6.0 

3.1 

3.9 

SC 

21.1 

±  7.1 

51.15 

34.0 

13.6 

34.8 

SCW 

5.6 

±  1.7 

2.78 

29.9 

3.6 

9.4 

1 ,000-SW 

270 

±  46.5 

2,163 

17.2 

184 

346 

Trait  parameters  across  four  populations 

CL 

4.7 

±  0.5 

0.30 

11.7 

3.7 

6.0 

CD 

4.2 

±  0.3 

0.11 

7.9 

3.1 

5.0 

SC 

37.8 

±  14.9 

222.22 

39.4 

13.6 

90.4 

SCW 

9.2 

±  3.7 

13.77 

40.3 

3.6 

23.1 

1 ,000-SW 

250 

±  49.7 

2,471.00 

19.8 

167.0 

354.0 

•  The  1,000-seed  weight  mean  was  238  +  47.4  g,  252  ± 
37.6  g,  238  ±  60.9  g,  and  270  ±  46.5  g,  respectively. 

Therefore,  the  southern  populations  were  similar  in 
cone  length  and  cone  diameter,  but  not  in  seeds  per  cone, 
seeds  per  cone  weight,  and  1,000-seed  weight.  On  the 
other  hand,  the  southern  populations  ranked  high  and  the 
northern  population  ranked  low  for  all  measured  traits. 

Very  high  variation  coefficients  were  found  within 
each  population  for  both  seeds  per  cone  and  seeds  per 
cone  weight,  but  middle  high  for  1,000-seed  weight  and 
low  for  cone  length  and  cone  diameter  traits. 

The  last  three  columns  of  table  3  give  information  con- 
cerning mean  squares,  variation  coefficients,  and  the 
range  of  variation  for  pop\ilation  traits. 

ACROSS-POPULATION 
PARAMETERS 

The  mean  values  and  ranges  of  the  traits  across  foiu* 
studied  populations — and  by  extrapolation — for  all  the 
Carpathian  stone  pine  were  as  follows  (table  3,  lower 
part): 

•  The  cone  length  mean  ranged  from  3.7  to  6.0  cm,  with 
a  mean  of  4.7  ±  0.5  cm,  and  the  cone  diameter  ranged 
from  3.1  to  5.0  cm,  with  a  mean  of  4.2  ±  0.3  cm.  According 
to  Contini  and  Lavarelo  (1992),  the  size  of  cones  from  the 
Alps  ranged  from  5.0  to  10.0  cm  in  length  and  from  4.0  to 
6.0  cm  in  diameter.  Therefore,  the  cone  size  from  the  Alps 
ranked  high  and  the  cones  from  the  Carpathians  ranked 
low. 

•  The  seeds  per  cone  ranged  from  13.6  to  90.4,  with  a 
mean  of  37.8  ±  14.9.  However,  the  seeds  per  cone  mean 
from  the  Alps  varied  between  46  and  164  with  a  mean 
of  93  seeds  (Rohmeder  and  Rohmeder  1955).  Thus,  the 
seeds  per  cone  from  the  Alps  ranked  high  and  the  seeds 
per  cone  from  the  Carpathians  ranked  low. 

•  The  seeds  per  cone  weight  mean  ranged  from  3.6  to 
23.1  g,  with  a  mean  of  9.2  ±  3.7  g. 

•  The  1,000-seed  weight  mean  ranged  from  167  to 
354  g,  with  a  mean  of  250  ±  49.7  g.  Consequently,  a  kilo- 
gram of  seed  from  the  Carpathians  could  include  from 
2,825  to  5,988  seeds,  with  a  mean  of  4,000  seeds.  Accord- 
ing to  Rohmeder  and  Rohmeder  (1955),  the  1,000-seed 
weight  mean  from  the  Alps  ranged  between  150  and 
350  g;  thus,  the  Carpathians  1,000-seed  weight  mean 
was  very  close  to  that  of  the  Alps. 

The  cone  length,  cone  diameter,  seeds  per  cone,  seeds 
per  cone  weight,  and  1,000-seed  weight  coefficients  ac- 
counted for  11.7  percent,  7.9  percent,  39.4  percent,  40.3 
percent,  and  19.8  percent  of  the  variation,  respectively 
(table  3,  lower  part). 

POLLINATION  COMPARISONS 

The  cone  and  seed  parameters  shown  in  table  4  came 
from  records  of  2,000  cones  from  80  open-pollinated  (OP) 
trees,  1,800  cones  from  a  10  x  10  full  diallel  mating  design 
for  cross-pollination  (CP),  and  200  cones  from  10  self- 
pollinated  (SP)  trees,  all  from  the  Gemenele  population. 
For  example,  the  cone  length  ranged  from: 


155 


3- 


2. 


Oi 

c 


300, 


250. 


200. 


.C150. 

Jioo 


o 
o 

°  50 


I 

/ 


I 

s 

; 

; 

s 
s 
s 
s 


1   2  3  4  X  X  X 


1  2  3  A  X  X  X 


2  3  A  X  X  X 


1  2  3  A  5f  X  f 


Figure  1 — Mean  performance  of  five  measured  traits  of  cones  and  seeds  from  open  pollination_ 
in  four  populations  (1 ,  2,  3,  4)  compared  to  controlled  cross-pollination  (f )  and  self-pollination  (^) 
(x=  average  across  1 ,  2,  3,  4  populations). 


4.1  to  5.8  cm,  with  a  mean  of  4.9  ±  0.4  cm  for  open- 
pollinated; 

3.5  to  6.0  cm,  with  a  mean  of  4.8  ±  0.6  cm  for  cross- 
pollinated; 

3.7  to  5.5  cm,  with  a  mean  of  4.7  ±  0.6  cm  for  self- 
pollinated. 

Variation  coefficients  (VC)  for  open-pollinated,  cross- 
pollinated,  and  self-pollinated  were  7.9  percent,  12.8  per- 
cent, and  12.1  percent,  respectively. 


Table  4 — Comparison  between  statistical  parameters  of  cones  and  seeds  in 
the  Gemenele  population,  according  to  the  pollination  type 


Type  of   Parameters 


Trait 

pollination^ 

k±a 

VC 

0 

CL 

OP 

4.9 

± 

0.4 

0.15 

7.9 

4.1 

5.8 

CP 

4.8 

± 

0.6 

0.37 

12.8 

3.5 

6.0 

SP 

4.7 

± 

0.6 

0.36 

12.1 

3.7 

5.5 

CD 

OP 

4.2 

+ 

0.3 

0.11 

7.9 

3.6 

5.0 

CP 

4.3 

+ 

0.4 

0.15 

9.1 

3.7 

5.4 

SP 

4.2 

± 

0.4 

0.16 

9.4 

3.6 

4.9 

SC 

OP 

40.9 

+ 

9.6 

91.31 

23.3 

22.5 

60.2 

CP 

62.0 

± 

16.5 

272.24 

26.6 

41.4 

96.4 

SP 

64.0 

+ 

12.9 

167.12 

20.3 

44.9 

85.7 

sew 

OP 

9.8 

+ 

2.7 

7.50 

28.0 

4.6 

18.1 

CP 

14.1 

± 

4.4 

19.56 

31.3 

7.0 

24.5 

SP 

13.8 

+ 

3.1 

9.66 

22.5 

9.5 

17.6 

1,000-SW  OP 

238.0 

± 

47.4 

2,243.00 

19.9 

170.0 

352.0 

CP 

188.0 

± 

31.3 

978.00 

16.6 

11.7 

26.5 

SP 

179.0 

± 

25.6 

656.94 

14.3 

14.7 

22.3 

'OP  =  open  pollination;  CP  =  controlled  cross-pollination  in  a  10  x  10  diallel;  SP  =  self 
pollination  of  10  parents. 


Similar  data  can  be  foimd  for  cone  diameter,  seeds  per 
cone,  seeds  per  cone  weight,  and  1,000-seed  weight  in 
table  4. 

CORRELATIONS 

Significant  (p  <  0.05)  and  highly  significant  (p  <  0.01) 
correlations  were  found  between  cone  length  and  cone  di- 
ameter, cone  length  and  seeds  per  cone,  cone  length  and 
seeds  per  cone  weight;  cone  diameter  and  seeds  per  cone, 
cone  diameter  and  seeds  per  cone  weight;  seeds  per  cone 
and  seeds  per  cone  weight.  No  significant  correlation 
was  found  between  1,000-seed  weight  and  any  other  trait 
(table  5).  These  strong  correlations  among  the  main  cone 
and  seed  traits  suggest  that  improvement  (quantitatively) 
of  seed  production  could  be  made  by  indirect  selection;  for 
example,  selection  for  cone  length,  as  an  easily  measur- 
able trait,  will  cause  an  increase  in  seeds  per  cone  and 
seeds  per  cone  weight,  and  consequently  in  seed  production. 

TYPE  OF  DISTRIBUTION 

The  distribution  frequencies  of  cone  length,  cone  diam- 
eter, seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight  were  very  close  to  normal  distribution  (fig.  2). 
According  to  genetic  theory  (Mather  and  Jinks  1977),  this 
pattern  of  distribution  is  specific  to  quantitative  traits. 
Such  traits  are  polygenically  controlled. 


156 


Table  5 — Phenotypic  correlations  among  Pinus  cembra  cone  and 
seed  traits  (Df  =  1 8) 


Trait 

CL 

CD 

SC 

sew 

1  nnn.sw 

CL 

1.000 

0.577- 

0.636- 

0.669- 

0.272 

CD 

1.000 

0.499* 

0.647- 

0.227 

SC 

1.000 

0.636- 

0.325 

sew 

1.000 

0.396 

1,000-SW 

1.000 

DISCUSSION 

Our  observations  in  the  main  Carpathian  populations 
indicated  that  tree,  locahty,  and  year  in  which  the  cones 
were  initiated  may  significantly  affect  cone  and  seed 
traits.  As  the  measurements  were  performed  on  cones 
collected  in  two  different  years,  the  comparisons  were 
perhaps  not  entirely  valid. 


The  seeds  per  cone  and  seeds  per  cone  weight  from  open- 
pollinated  cones  ranked  lower  than  the  same  parameters 
from  both  cross-pollinated  and  self-pollinated  cones;  but, 
surprisingly,  1,000-seed  weight  from  open-pollinated 
cones  ranked  higher  than  1,000-seed  weight  from  cross- 
pollinated  cones.  The  lack  of  significant  correlation 
between  1,000-seed  weight  and  seeds  per  cone  weight 
(table  5)  partially  explains  this  unexpected  resvilt.  Also, 
these  differences  could  be  attributable  to  the  biological 
and  climatic  factors  (temperature,  moistm-e,  and  wind) 
that  occurred  in  1990  and  1991. 

The  cone  length  and  cone  diameter  means  from  self- 
pollinated  cones  were  similar  to  cone  length  and  cone 
diameter  means  from  both  open-pollinated  and  cross- 
pollinated  cones,  but  the  seeds  per  cone  and  seeds  per 
cone  weight  means  from  self-pollinated  cones  were  similar 
to  seeds  per  cone  and  seeds  per  cone  weight  means  from 
cross-pollinated  cones  and  ranked  higher  them  open- 
pollinated  cones.  Consequently,  stone  pine  was  found 


£  20 


30 


^  20  - 


O 


£  10 


330 


3.7    3.9     4.2    4.5    4.8    5.1     5.4    5.7  6.0 

Cone  Length  (cm) 


3.1    3.4     3.7    4.0    4.3   4.6  4.9 

Cone  Diameter  (cm) 


^  20 


0> 


S>  10 


10    20     30     40     50     60     70     80  90 

Seeds  per  Cone  (No.) 


30  - 


^   20  ■ 


o 


2>  10 


4     7     10    13    16    19  22 


>g  20- 


0) 


0) 


150  190    230   270  310  350    390  430 


Seeds  per  Cone  Weight  (g)  1 ,000-Seed  Weight  (g) 

Figure  2— Frequency  distribution  in  the  Pinus  cembra  populations  evaluated  for  five  traits. 


157 


to  be  highly  self-fertile.  This  led  to  the  conclusion  that 
a  very  high  proportion  of  the  seeds  produced  by  wind  pol- 
lination would  be  selfs;  this  is  an  undesirable  characteris- 
tic since  selfed  seed  produces  slower  growing  trees  and 
the  seedlings  have  a  lower  survival  rate. 

Finally,  it  should  be  stressed  that  although  the  stone 
pine  cone  diameter,  cone  length,  and  seeds  per  cone 
means  from  the  Alps  exceeded  the  same  trait  means  from 
the  Carpathians,  the  1,000-cone  weight  mean  from  the 
Carpathians  was  similar  to  that  from  the  Alps. 

CONCLUSIONS 

Southern  populations  were  similar  in  cone  length  and 
cone  diameter  but  not  in  seeds  per  cone,  seeds  per  cone 
weight,  and  1,000-seed  weight.  Southern  populations 
ranked  high  and  the  northern  ones  ranked  low  in  all 
measured  traits. 

Very  high  variation  in  both  seeds  per  cone  and  seeds 
per  cone  weight  was  found  within  each  population,  but 
variation  in  1,000-seed  weight  was  moderate,  and  it  was 
low  in  cone  length  and  cone  diameter  traits. 

Cone  length  and  cone  diameter  from  open-pollinated, 
cross-pollinated,  and  self-pollinated  cones  were  similar, 
while  seeds  per  cone,  seeds  per  cone  weight,  and  1,000- 
seed  weight  were  not. 

The  strong  correlations  among  the  main  cone  and  seed 
traits  suggest  that  genetic  improvement  in  seed  produc- 
tion could  be  attained  by  indirect  selection. 

All  measured  traits  displayed  a  continuous  variation, 
suggesting  polygenic  control. 

ACKNOWLEDGMENTS 

Gratitude  is  expressed  to  Dr.  H.  Kriebel  from  Ohio 
State  University  for  the  revision  of  the  English  version 
of  the  paper  and  for  the  useful  suggestions. 

REFERENCES 

Beldie,  A.  1941.  Observatii  asupra  vegetatiei  lemnoase 
din  Muntii  Bucegi.  Analele  ICEF,  Seria  1,  vol.  6:  39-43. 

Blada,  I.  1982.  Relative  blister  rust  resistance  of  native 
and  introduced  white  pines  in  Romania.  In:  Heybroek,  H.; 
[and  others],  eds.  Resistance  to  diseases  and  pests  in 
forest  trees.  PUDOC,  Wageningen,  The  Netherlands: 
415-416. 

Blada,  I.  1987.  Genetic  resistance  to  Cronartium  ribicola 
and  height  growth  in  some  five  needle  pines  and  of 


some  inter-specific  hybrids.  Bucharest:  Academy  of 
Agricultural  and  Forestry  Sciences.  146  p.  Thesis. 

Blada,  I.  1990a.  Breeding  Pinus  cemhra  by  intra-  and 
inter-specific  crosses.  For.  Res.  Inst.,  Bucharest,  Re- 
search Plan,  No.  7/1990. 

Blada,  I.  1990b.  Genetic  variability  of  some  traits  in  two 
Pinus  cembra  natural  populations.  In:  Garrett,  P.  W., 
ed.  Proceedings  of  a  s3miposium  on  white  pines  prov- 
enances and  breeding.  Gen.  Tech.  Rep.  NE-155.  Radnor, 
PA:  U.S.  Department  of  Agriculture,  Forest  Service, 
Northeastern  Forest  Experiment  Station:  56-68. 

Blada,  I.  1992a.  Pinus  cembra  and  its  inter-specific  hy- 
brids. Paper  presented  at  the  lUFRO  Centennial  Meet- 
ing, Berlin-Eberswalde,  August  31-September  5,  1992. 

Blada,  I.  1992b.  Diallel  cross  in  Pinus  cembra.  I.  Variation 
in  cone  and  seed  traits.  [In  preparation]. 

Contini,  L.;  Lavarelo,  Y.  1982.  Le  pin  cembro:  repartition, 
ecologie,  sylviculture  et  production.  Paris:  INRA,  ISBN. 
197  p. 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  dis- 
tribution of  the  pines  of  the  world.  IVIisc.  Publ.  991. 
Washington,  DC:  U.S.  Department  of  Agriculture,  For- 
est Service.  97  p. 

Griffing,  B.  1956.  Concept  of  general  and  specific  com- 
bining ability  in  relation  to  diallel  crossing  systems. 
Australian  J.  Sci.  9:  465-493. 

Gubesch,  L.  1971.  Raspandirea  relictului  glaciar  zambrul 
(Pinus  cembra)  pe  versantii  sudici  ai  unor  masive  din 
Calimani.  Ocrotirea  Naturii,  15,  2:  149-159. 

Holzer,  K.  1972.  Intrinsic  qualities  and  growth  potential 
of  Pinus  cembra  and  P.  peuce  in  Europe.  In:  Bingham, 
R.  T.;  [and  others],  eds.  Biology  of  rust  resistance  in  for- 
est trees.  Misc.  Publ.  Washington,  DC:  U.S.  Department 
of  Agriculture,  Forest  Service:  99-110. 

Holzer,  K.  1975.  Genetics  of  Pinus  cembra  L.  Anales 
Forestales,  Zagreb.  6/5:  139-158. 

Mather,  K.;  Jinks,  J.  L.  1977.  Introduction  to  biometrical 
genetics.  Chapman  and  Hall,  Ltd.  233  p. 

Moser,  L.  1960.  Verbreitung  und  Bedentung  der  Zirbe 
im  italienishcen  Alpengebiet.  Jb.  Ver.  Z.  Schutz 
Alpenpfl.u.-tiere.  25:  16-21. 

Oarcea,  Z.  1966.  Contributii  la  cunoasterea  raspandirii  si 
vegetarii  pinului  cembra  in  Retezat.  Revista  Pad.  9: 
495-497. 

Rohmeder,  E.;  Rohmeder,  M.  1955.  Untersuchungen  uber 
das  Samentragen  und  Keimen  der  Zirbelkiefer  (Pinus 
cembra)  in  den  Bayrischen  Alpen.  Allg.  Forstzeitschrift. 
10:  83. 

Tataranu,  I.  D.;  Costea  C.  1952.  Un  arbore  de  interes 
forestier:  Pinus  cembra.  Revista  Padurilor.  11:  3-14. 


158 


ROLE  OF  NUTCRACKERS  ON  SEED 
DISPERSAL  AND  ESTABLISHMENT  OF 
PINUS  PUMILA  AND  P.  PENTAPHYLLA 

Mitsuhiro  Hayashida 


AbBtract— Nutcrackers  (Nucifraga  caryocatactes)  harvest  almost 
all  Pinus  pumila  Kegel  cones  and  cache  the  seeds  in  the  soil.  Most 
P.  pentaphylla  'M.ayr  seeds  scatter  after  the  cones  open,  while  nut- 
crackers collect  the  seeds  still  held  in  the  cones  and  then  cache  the 
seeds.  Many  seedlings  of  P.  pumila  and  P.  pentaphylla  are  often 
found  in  clusters  consisting  of  several  individuals  of  the  same  age. 
Almost  all  P.  pumila  seedlings  and  clusters  of  P.  pentaphylla 
seedlings  on  barren  slopes  likely  originate  from  seeds  cached  by 
nutcrackers. 


Stone  pines  (subsection  Cembrae  in  Pinus)  have  large 
wingless  seeds  that  are  dispersed  by  caching  of  vertebrates 
(Hayashida  1989a;  Hutchins  and  Lanner  1982;  Mattes 
1982;  Tomback  1982).  Nutcrackers  {Nucifraga  spp.)  are 
the  main  dispersal  agents  of  seeds  of  these  pines  (Lanner 
1989).  The  nutcracker  also  caches  other  large  wingless 
seeds  or  short-winged  seeds  in  section  Strobus  (Lanner 
and  Vander  Wall  1980;  Tomback  1990;  Vander  Wall  and 
Balda  1977). 

Pinus  pumila  Regel  (Japanese  stone  pine)  is  a  dwarf  pine. 
Its  range  extends  through  Japan  and  Korea  into  Siberia, 
Kamchatka  (Critchfield  and  Little  1966).  On  most  of  the 
high  mountains  in  Japan,  a  vegetational  zone  dominated 
by  P.  pumila  occurs  above  the  forest  limit.  This  area  is 
called  the  "Pinus  pumila  zone"  by  Japanese  ecologists.  The 
regeneration  of  P.  pumila  thickets  is  mainly  due  to  a  veg- 
etative regeneration  by  adventitious  roots  (Okitsu  and  Ito 
1983).  Large  wingless  seeds  of  P.  pumila  are  known  to  be 
dispersed  by  nutcrackers  (Saito  1983;  Turcek  and  Kelso 
1964),  but  a  detailed  observation  for  seed  dispersal  and 
caching  behavior  is  needed. 

Pinus  parviflora  Sieb.  et  Zucc.  (Japanese  white  pine)  is 
distributed  throughout  the  Japanese  islands  (Critchfield 
and  Little  1966).  This  species  consists  of  two  geographical 
varieties,  which  intergrade  in  central  Japan  (Hayashi  1954). 
The  northern  variety  is  sometimes  called  P.  pentaphylla 
Mayr;  the  southern  P.  himekomatsu  Miyabe  and  Kudo  or 
a  variety  of  the  former.  Since  the  seed  wings  of  P.  penta- 
phylla are  longer  than  those  of  P.  himekomatsu  (Ishii  1968), 
I  distinguished  between  the  two  pines.  There  are  few  eco- 
logical studies  of  these  pines  except  for  community  struc- 
ture (Tatewaki  and  others  1960;  Yoshioka  and  Saito  1962). 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Mitsuhiro  Hayashida  is  Associate  Professor,  Department  of  Bioenviron- 
ment,  Faculty  of  Agriculture,  Yamagata  University,  Wakaba,  Tsuruoka 
997,  Japan. 


Pinus  pumila  and  P.  pentaphylla  are  in  the  subgenus 
Strobus  and  are  so  similar  that  they  naturally  hybridize 
(P.  hakkodensis  Makino)  (Ishii  1968);  but  these  pines  differ 
in  seed,  cone,  and  other  characteristics.  The  objectives  of 
this  paper  are  to  describe  the  processes  from  seed  dispersal 
to  seedling  estabUshment  of  P.  pumila  and  P.  pentaphylla 
and  to  examine  the  role  of  nutcrackers  in  their  regeneration. 

STUDY  AREA 

The  study  was  conducted  on  Mount  Apoi  (811  m  above  sea 
level),  which  is  located  in  south-central  Hokkaido,  Japan 
(42°6'  N.,  143°2'  E.).  It  is  composed  of  ultramafic  rocks 
(dunite,  IherzoHte,  etc.)  (Niida  1984),  and  is  covered  with 
coniferous  forests.  The  P.  pumila  zone  extends  from  eleva- 
tions of  500  m  to  800  m.  Below  500  m,  P.  pentaphylla  for- 
est occupies  the  rocky  slopes  and  ridges,  while  the  greater 
part  of  the  mountain  is  covered  by  coniferous  forests  domi- 
nated by  Picea  glehnii,  Pinus  pentaphylla,  and  Abies 
sachalinensis  (Hayashida  1989b). 

METHODS 

The  number  and  distribution  of  2-year  cones  were  mapped 
to  determine  rates  of  seed  predation  by  vertebrates  and  in- 
sects. These  cones  were  counted  from  July  to  October  in 
1984-87.  The  observations  on  seed  harvesting,  transport- 
ing, caching,  and  retrieving  behavior  in  vertebrates  were 
made  in  1984-87  using  binoculars  and  a  telescope. 

Many  seedlings  of  P.  pumila  and  P.  pentaphylla  were 
found  on  barren  slopes  along  mountain  trails.  These  seed- 
lings were  often  found  in  clusters,  which  consisted  of  sev- 
eral seedhngs  of  the  same  age.  To  obtain  the  frequency 
distribution  of  number  of  individuals  per  cluster  in  both 
pines,  all  seedlings  on  barren  slopes  along  mountain  trails 
were  recorded. 

SEED  AND  CONE  MORPHOLOGY 

Pinus  pumila  has  wingless  seeds;  P.  pentaphylla  seeds 
have  wings  that  are  seed  length  (table  1).  Seeds  of  both 
pines  have  the  same  thick  coats.  Pinus  pumila  seeds  are 
significantly  heavier  than  those  of  P.  pentaphylla,  though 
the  seeds  of  P.  pentaphylla  are  larger. 

Pinus  pumila  cones  do  not  open  when  they  ripen,  but 
P.  pentaphylla  cones  open  at  maturity  and  release  seeds 
to  fall  free.  Pinus  pumila  cones  weigh  less  than  P.  penta- 
phylla cones,  but  contain  more  seeds. 


159 


Table  1 — Seed  and  cone  characteristics  of  Pinuspumila  and  P.  pentaphylla 


Pinus  pumila    P.  pentaphylla 


MAAQIirOITIAnt 

d  1  Id  It 

Moan  +  Qn 

M 

Moon  4-  QD 

Kl 

n 

P< 

Seed  length  (mm) 

7.700  ±  0.600 

105 

10.000  ±  0.700 

105 

0.001 

Dry  weight  (g) 

0.100  ±  0.014 

36 

0.068  ±  0.014 

20 

0.001 

Seed  coat  thickness  (mm) 

0.390  ±  0.040 

20 

0.390  ±  0.040 

20 

NS 

Length  of  wings  (mm) 

wingless 

10.500  +  1.600 

105 

Cone  mass  (g) 

11.00013.300 

32 

24.800  ±  7.000 

27 

0.001 

Number  of  seeds  per  cone 

42.800  ±12.900 

45 

28.900  ±9.100 

38 

0.001 

Dehiscence 

not  open 

open 

'Mann-Whitney  U  test. 


SEED  DISPERSAL  OF  P.  PUMILA 

Figure  1  shows  the  seasonal  disappearance  of  2-year 
cones  of  P.  pumila  in  1985-88.  All  cones  disappeared  from 
trees  by  mid-October  every  year.  I  observed  nutcrackers 
(Nucifraga  caryocatactes),  squirrels  (Sciurus  vulgaris  ori- 
entis),  and  chipmunks  (Tamias  sibiricus  lineatus)  harvest- 
ing cones.  Five  species  of  vertebrates  harvested  and  car- 
ried seeds  to  caches:  nutcracker,  vsiried  tit  (Parus  varius), 
nuthatch  (Sitta  europaea),  squirrel,  and  chipmunk.  The 
nutcracker  was  the  most  frequently  observed  vertebrate 
carr3dng  P.  pumila  seeds. 

Nutcrackers  usually  harvest  seeds  by  removing  the 
cones  and  flying  with  them  to  perches  where  they  extract 
the  seeds  and  hold  them  in  their  sublingual  pouches.  Var- 
ied tits  and  nuthatches  were  able  to  take  P.  pumila  seeds 
exposed  by  nutcracker  foraging  and  eat  them.  They  were 
seen  caching  pine  seeds  on  a  tree  or  in  the  soil.  Squirrels 
visited  the  P.  pumila  scrub  from  a  coniferous  forest  to  for- 
age on  pine  seeds.  They  often  carried  a  whole  cone  to  cache 
it  in  scatter-hoarding.  Chipmunks  harvest  pine  seeds  in 
their  cheek  pouches.  They  carried  about  50  seeds  on  a  trip 
and  cached  them  in  the  soil. 

These  results  indicate  that  nutcracker  is  the  main  seed 
disperser  of  P.  pumila. 


(0 

E 
« 
oc 


Jul. 

Aug. 

Sep. 

Oct. 

Figure  1 — Retention  of  mature  cones  of  Pinus  pumila 
on  trees  (in  years  1985-88)  as  related  to  season. 


SEED  DISPERSAL  OF 
P.  PENTAPHYLLA 

Cones  taken  by  vertebrates  or  damsiged  by  insects 
(cone-boring  lepidopteran  larvae)  accounted  for  less  than 
20  percent  of  the  total.  Usually,  nearly  90  percent  of  the 
cones  open  on  the  trees  from  mid-September  through  early 
October.  After  opening,  about  80  percent  of  the  seeds  im- 
mediately scatter  to  the  ground,  due  to  the  fact  that  77 
percent  of  P.  pentaphylla  cones  were  oriented  downward 
(Hayashida  1989b). 

Nutcrackers,  varied  tits,  and  nuthatches  harvested  and 
cached  seeds  still  held  in  the  upward-oriented  open  cones. 
In  the  years  when  P.  pumila  crop  sizes  were  small,  nut- 
crackers harvested  P.  pentaphylla  seeds  more  frequently. 
They  extracted  the  several  seeds  from  an  open  cone,  peeled 
off  the  seed  wings,  and  put  seeds  into  their  sublingual 
pouches.  They  moved  rapidly  around  a  tree  and  went  from 
tree  to  tree. 

NUTCRACKER  CACHING  BEHAVIOR 

Nutcrackers  began  to  cache  P.  pumila  seeds  in  early 
August  when  the  seeds  were  mature,  and  they  continued 
to  harvest  seeds  until  the  seed  crop  was  depleted.  They 
cached  the  seeds  in  the  soil  at  a  depth  of  2  to  3  cm.  Cache 
size  (the  number  of  seeds  per  cache)  ranged  from  1  to  51. 
Nutcrackers  cached  P.  pumila  seeds  mainly  in  coniferous 
forests. 

In  mid-September,  P.  pentaphylla  cones  began  to  open, 
then  nutcrackers  began  to  harvest  seeds  from  the  opening 
cones.  Nutcrackers  scatter-hoarded  almost  all  seeds  in  the 
soil  (97  percent)  except  for  two  caches  containing  one  and 
two  seeds  that  were  established  in  the  bark  on  fallen  trees. 
Cache  size  ranged  from  1  to  40  seeds.  Seventy-eight  caches 
were  recorded  in  coniferous  forests  and  only  one  cache  in 
the  Pinus  pumila  zone. 

CLUMPS  OF  SEEDLINGS 

Eighty-four  percent  of  the  occurrence  of  P.  pumila  seed- 
lings were  in  groups  of  two  or  more  trees  (table  2).  These 
seedlings  were  1  km  in  distance  from  the  nearest  seed  tree. 


160 


Table  2 — Number  of  individuals  per  clump  of  Pinus  pumila  and  P.  pentaphylla  on 
barren  slopes  along  mountain  trails 


Number  of  clumoQ 

Numbor  of  inHiviclijalQ 

(percent) 

per  clump 

Species 

Single       Clump  (2<) 

Maximum       Mean  ±  SD 

Pinus  pumila 

9(16)  47(84) 

26              5.6  ±5.3 

P.  pentaphylla 

137(72)  54(28) 

25              2.6  ±  3.8 

Twenty-eight  percent  of  P.  pentaphylla  seedlings  were 
growing  in  groups  of  two  or  more  trees.  There  were  many 
seed  trees  of  P.  pentaphylla  along  the  moimtain  trails.  Sin- 
gle seedlings  might  originate  from  naturally  scattered  fallen 
seeds,  and  the  seedlings  in  cliunps  probably  originated  from 
caches  by  nutcrackers. 

DISCUSSION 

All  mature  cones  of  P.  pumila  are  harvested  by  verte- 
brates every  year  (fig.  1).  My  observations  indicate  that 
almost  all  of  the  transported  seeds  were  carried  by  nut- 
crackers. Nutcrackers  cached  seeds  in  the  soil  at  a  depth 
of  2  to  3  cm.  It  is  shallow  enough  to  permit  seedling  estab- 
lishment. Therefore,  seed  dispersal  of  P.  pumila  is  heavily 
dependent  on  seed  caching  by  nutcrackers. 

Nutcrackers  carried  and  cached  P.  pumila  seeds  mainly 
into  coniferous  forests.  In  general,  P.  pumila  seedhngs  can- 
not grow  and  establish  in  coniferous  forests  even  if  cached 
seeds  survive  and  germinate.  However,  several  P.  pumila 
trees  were  found  that  developed  and  matured  on  the  rocky 
ridges  or  barren  slopes  in  coniferous  forests.  These  facts 
may  indicate  that  the  nutcracker  is  not  an  efficient  seed 
disperser  of  P.  pumila  under  the  present  conditions.  Pinus 
pumila  adapted  its  specialized  tree  form  and  other  features 
to  the  habitat  that  is  characterized  by  strong  winds  and 
heavy  snow  accumulation  in  winter.  Thus,  the  P.  pumila 
zone  is  developed  on  deforested  areas  in  the  boreal  subal- 
pine  belt.  In  Europe,  nutcrackers  cached  P.  cembra  seeds 
in  coniferous  stands  and  sometimes  cached  them  above 
timberline  (Mattes  1982).  Seed  dispersal  over  wide  areas 
by  nutcrackers  probably  played  an  important  role  in  ex- 
panding the  range  of  P.  pumila. 

Pinus  pentaphylla  has  dehiscent  cones  and  winged  seeds. 
Most  of  the  seeds  scattered  when  cones  opened.  Nutcrack- 
ers often  harvested  and  cached  seeds  still  held  in  the  up- 
ward open  cones.  About  90  percent  of  P.  pentaphylla  trees 
found  in  coniferous  forests  were  individual  single  trees,  but 
on  the  ecotone  between  coniferous  forest  and  the  Pinus 
pumila  zone,  half  of  the  occurrences  of  trees  were  in  clumps 
of  two  or  more  individuals  (Hayashida  1989b).  About  30 
percent  of  the  seedlings  were  growing  in  clumps  on  barren 
slopes  in  the  coniferous  forest  zone.  These  seedlings  in 
climips  probably  originated  from  seeds  cached  by  nutcrack- 
ers. Seed  dispersal  by  nutcrackers  enables  P.  pentaphylla 
to  invade  areas  that  are  inaccessible  to  other  coniferous 
trees. 

The  northern  variety,  P.  pentaphylla,  has  longer  seed 
wings  than  the  southern  variety,  P.  himekomatsu  (Ishii 
1968).  It  is  possible  that  these  two  varieties  have  different 


morphological  features  and  seed-dispersal  syndromes.  If 
so,  why?  I  would  study  these  problems  with  an  ecological 
approach. 

ACKNOWLEDGMENTS 

I  thank  Urakawa  Local  Forest  Management  Office, 
Hokkaido,  and  Oji  Forestry  and  Landscaping  Co.  for  per- 
mission to  work  in  the  area  and  for  information  on  the 
area. 

REFERENCES 

Critchfield,  W.  B.;  Little,  E.  L.,  Jr.  1966.  Geographic  distri- 
bution of  pine  of  the  world.  Misc.  Publ.  991.  Washington, 
DC:  U.S.  Department  of  Agriculture,  Forest  Service.  97  p. 

Hayashi,  Y.  1954.  The  natural  distribution  of  important 
trees,  indigenous  to  Japan.  Conifers  Rep.  3.  Bulletin  of 
the  Government  Forestry  Experiment  Station.  75:  1-173. 
[In  Japanese  with  English  svmimary]. 

Hayashida,  M.  1989a.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecology 
and  Management.  28:  115-129. 

Hayashida,  M.  1989b.  Seed  dispersal  and  regeneration 
pattern  of  Pinus  parviflora  var.  pentaphylla  on  Mt.  Apoi 
in  Hokkaido.  Research  Bulletins  of  the  College  Experi- 
ment Forests,  Facvdty  of  Agriculture,  Hokkaido  Univer- 
sity. 46(1):  177-190.  [In  Japanese  with  English  stmimary]. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Ishii,  S.  1968.  The  basic  forestry  study  on  genus  Pinus, 
especially  its  taxonomic  and  geographical  consideration. 
Journal  of  the  Faculty  of  Agriculture,  Kochi  University. 
19:  1-114.  [In  Japanese  with  English  svunmary]. 

Lanner,  R.  M.  1989.  Biology,  taxonomy,  evolution,  and 
geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposiimi  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mouintain  resource;  1989  March 
29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service,  In- 
termountain  Research  Station:  14-24. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Journal  of  Forestry. 
78:  637-639. 

Mattes,  H.  1982.  Die  Lebensbemeinschaft  von  Tannenhaher, 
Nucifraga  caryocatactes  (L.),  und  Arve,  Pinus  cembra  L., 
und  ihre  forstliche  Bedeutiing  in  der  oberen  Gebirgs- 
waltstufe,  Berichte.  241.  74  p. 


161 


Niida,  K.  1984.  Petrology  of  the  Horoman  utramafic  rocks 
in  the  Hidaka  Metamorphic  Belt,  Hokkaido,  Japan.  Jour- 
nal of  the  Faculty  of  Science,  Hokkaido  University,  Ser. 
IV.  21:  197-250. 

Okitsu,  S.;  Ito,  K.  1984.  Vegetation  d3niamics  of  the  Siberian 
dwarf  pine  (Pinus  pumila  Kegel)  in  the  Taisetsu  moun- 
tain range  Hokkaido,  Japan.  Vegetatio.  58:  105-113. 

Saito,  S.  1983.  Caching  of  Japanese  stone  pine  seeds  by 
nutcrackers  at  the  Shiretoko  Peninsula,  Hokkaido.  Tori. 
32:  13-20.  [In  Japanese  with  English  summary]. 

Tatewaki,  M.;  Tsujii,  T.;  Kawano,  S.  1960.  The  community 
structure  and  distribution  of  Pinus  parvifLora  forest  in 
Hokkaido.  Japanese  Journal  of  Ecology.  10:  120-123.  [In 
Japanese  with  English  summary]. 


Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds  by 
Clark's  nutcracker:  a  mutualism  hypothesis.  Journal  of 
Animal  Ecology.  51:  451-467. 

Tomback,  D.  F.  1990.  The  evolution  of  bird-dispersed 
pines.  Evolutionary  Ecology.  4:  185-219. 

Turcek,  F.  J.;  Kelso,  L.  1968.  Ecological  aspects  of  food 
transportation  and  storage  in  the  Corvidae.  Communi- 
cations in  Behavioral  Biography,  Al:  277-297. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of  the 
Clark's  nutcracker  and  the  pinon  pine  for  efficient  seed 
harvest  £md  dispersal.  Ecological  Monographs.  47:  89-111. 

Yoshioka,  K;  Saito,  K.  1962.  Differences  in  the  distribution 
of  forest  communities  due  to  topography  of  the  habitats 
in  Okunikkawa,  Miyagi  Prefecture.  Ecological  Review. 
15:  213-220. 


162 


ROLE  OF  VARIOUS  ANIMALS  IN 
DISPERSAL  AND  ESTABLISHMENT 
OF  WHITEBARK  PINE  IN  THE  ROCKY 
MOUNTAINS,  U.S.A. 

Harry  E.  Hutchins 


Abstract — The  dispersal  and  establishment  of  whitebark  pine 
has  been  attributed  to  Clark's  nutcracker.  But  what  is  the  role 
of  other  animals  who  forage  on  whitebark  pine  seed?  This  study 
looks  at  the  role  of  the  diurnal  birds  and  mammals  that  forage 
on  whitebark  pine  cones  and  seed  in  western  Wyoming.  A  total 
of  1,005  cones  were  carefully  observed  in  both  contiguous  "forest" 
stands  and  open-grown  "meadow"  trees  in  early  November.  Red 
squirrels  harvested  most  of  the  whitebark  pine  seed  in  forested 
plots  while  Clark's  nutcracker  removed  most  of  the  seed  from  the 
meadow  trees.  Only  a  small  percentage  of  seed  went  to  ground 
squirrels,  chipmunks,  Stellar's  jays,  ravens,  etc.  The  foraging 
behavior  as  well  as  the  importance  of  these  animal  species  to 
whitebark  pine  regeneration  is  discussed. 


Many  animal  species  have  been  observed  foraging  on 
the  nutritious  seeds  of  whitebark  pine  (Pinus  albicaulis). 
But  exactly  which  of  these  animals  plays  a  role  in  seed- 
ling establishment  and  to  what  degree?  Which  animals 
are  seed  predators?  Can  whitebark  pine  seed  germinate 
from  that  seed  which  is  dropped  through  animal  foraging 
accidents?  Do  animals  harvest  all  the  seed  in  the  fall  or 
can  cones  fall  to  the  ground  and  seed  germinate  as  the 
cone  decays?  This  paper  will  address  these  questions. 

The  role  of  various  animals  in  the  dispersal  and  estab- 
lishment of  whitebark  pine  was  examined  by  observing 
1,005  cones  and  determining  where  and  how  the  seeds 
were  dispersed.  By  primarily  observing  the  behavior  of 
diurnal  birds  and  mammals  in  the  subalpine  canopy,  I 
was  able  to  determine  the  degree  of  influence  each  species 
has  on  seed  dispersal  and  seedling  establishment. 

METHODS 

The  study  was  primarily  conducted  at  the  Squaw  Basin- 
Togwotee  Pass  area  (Bridger-Teton  National  Forest;  fig.  1). 
This  area  of  high-elevation  meadows  offered  distant  views 
of  bird  activity,  whitebark  pine  growing  in  both  contigu- 
ous forest  and  open-grown  situations,  and  an  abundant 
cone  crop  for  whitebark  pine  diuing  1980.  Besides  white- 
bark pine,  the  forest  stands  were  composed  of  Engelmann 
spruce  (Picea  engelmannii),  subalpine  fir  {Abies  lasio- 
carpa),  and  a  minor  amoimt  of  lodgepole  pine  (Pinus  con- 
torta).  The  moraine  meadow  ridges  were  pioneered  by 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Harry  E.  Hutchins  is  Natural  Resource  Instructor,  Itasca  Community 
College,  Grand  Rapids,  MN. 


lone  whitebark  pine.  This  provided  an  opportimity  to 
study  two  different  types  of  commimities — "forest"  and 
"meadow"  (fig.  2). 

Observations  of  animal  activity  were  also  made  at 
Mount  Washburn  (2,680  to  3,140  m,  Yellowstone  National 
Park,  WY)  and  Surprise  Lake  (2,960  m.  Grand  Teton  Na- 
tional Park,  WY).  These  areas  were  similar  to  the  Squaw 
Basin  site  except  they  lacked  open-grown,  cone-bearing 
trees.  Details  on  the  study  sites  can  be  foxmd  in  Hutchins 
and  Lanner  (1982)  and  Vander  Wall  and  Hutchins  (1983). 

Several  cone-bearing  whitebark  pines  were  chosen  for 
observation  in  both  continuous  forest  in  which  red  squir- 
rels {Tamiasciurus  hudsonicus)  were  present  and  in  the 
open  meadow,  at  least  several  hundred  meters  from  the 
nearest  forest  edge  (these  sites  lacked  squirrels).  In  each 
of  these  two  types  of  sites,  1,005  whitebark  pine  cones 
were  observed  during  the  period  Jvdy  3  to  November  2, 
1980.  The  cones  were  scattered  among  several  mature 
trees  and  were  counted  at  1-  to  2-week  intervals  using 
a  15-25  X  60-mm  spotting  scope.  Coimts  were  made  by 
standing  in  a  specific  marked  location  and  "mapping"  the 
cones  on  clear  acetate.  Changes  in  the  "cone  map"  at  each 
observation  were  recorded,  including  the  partial  removal 
of  a  cone.  Cone  count  data  were  converted  into  seeds  us- 
ing an  empirically  derived  value  for  the  mean  number  of 
seeds  per  cone  (50.4  seeds/cone,  Hutchins  1982).  Partially 
consimied  cones  were  tallied  by  estimating  from  the  ground 
the  percentage  of  seed  remaining  as  described  in  Hutchins 
and  Lanner  (1982).  The  seed  harvest  data  were  then  plot- 
ted against  ciunulative  time  for  both  forest  and  meadow 
sites.  At  each  observation  date  samples  of  seeds  were  col- 
lected to  determine  maturity  and  condition.  The  numbers 
of  filled,  discolored,  insect-attacked,  and  second-year- 
aborted  ovules  were  tallied.  Mean  dry  weight  of  shelled 
seeds,  seed  coat  thickness,  and  caloric  content  of  shelled 
seeds  were  all  obtained.  Germination  tests  were  also  con- 
ducted with  seeds  collected  at  10  collection  dates  during 
the  1980  field  season. 

Seedfall  (caused  by  animal  foraging)  below  the  tree  was 
estimated  by  placing  five,  1-m^  vnre  mesh  seed  traps  ran- 
domly below  trees.  The  tops  of  the  seed  traps  were  de- 
signed to  let  seed  fall  through  but  to  exclude  rodents. 
Data  on  the  nimiber  of  cones  in  the  tree,  the  area  of  the 
tree  crown,  and  the  number  of  seeds  falling  per  square 
meter  were  used  to  estimate  the  magnitude  of  seed  fall. 

Predation  on  seed  caches  of  whitebark  pine  was  also 
studied  in  1979  and  1980.  I  simulated  three  types  of 
caches:  (1)  seed  that  lands  on  the  soil  surface  from 
foraging  accidents;  (2)  seed  cached  at  a  depth  of  3  cm 


163 


MONTANA 


WYOMING 


GRAND  TEfoN 
NAT.'  PARk 

I    i  ® 

Surprise  Lake  ^ 

i  A 


SHOSHONE 
NATIONAL 
FOREST 


LEGEND 


—  PAVED  ROAD 
®      STUDY  SITE 

KILOMETERS 


WYOMING 


simulating  a  Clark's  nutcracker  (Nucifraga  columbiana) 
cache;  and  (3)  seed  cached  7  cm  deep,  which  simulates 
the  most  shallow  red  squirrel  seed  cache.  Each  individual 
cache  contained  10  seeds. 

Animal  time  budget  data  were  collected  for  the  diurnal 
species  found  foraging  in  the  whitebark  pine  ecosystem. 
Both  quantitative  and  qualitative  observations  were  made 
on  the  various  activities  and  behaviors  of  the  animals;  the 
methods  are  detailed  in  Hutchins  and  Lanner  (1982). 

SEED  DEVELOPMENT 

Hutchins  and  Lanner  (1982)  monitored  seed  develop- 
ment and  found  mean  seed  coat  thickness  and  seed  weight 
to  be  significantly  greater  during  the  August  31  to 
November  2  collecting  period  than  prior  to  these  dates. 
Clark's  nutcrackers  were  unable  to  extract  whole  seed 
from  a  cone  ;intil  August  13  due  to  the  thin,  fragile  seed 
coats.  The  characteristic  shell  fragments  of  an  unripened 
"nutcracker  cone"  were  left  as  evidence  of  their  foraging 
attempts.  Thus,  nutcrackers  were  unable  to  cache  whole 
seed  until  after  this  date. 

By  September  7-10,  the  cones  had  dried  and  turned  a 
dull  brown  from  their  previous  moist,  pitch-filled,  purple 
color.  Whitebark  pine  cones  are  often  referred  to  as  inde- 
hiscent;  however,  about  25  percent  of  the  cones  collected 
after  September  7  parted  their  scales  slightly  (4-8  mm; 
n  =  141).  This  still  was  not  enough  of  an  opening  to  allow 
the  seeds  to  fall  out  on  their  own  accord. 


Figure  1 — Locations  of  three  study  sites  in  Wyoming. 


164 


SEED  CROP  DEPLETION 

Figure  3  (from  Hutchins  and  Lanner  1982)  shows  seed 
crop  depletion  followed  a  logistic  curve  (r^  =  0.99  in  the 
forest  and    =  0.96  in  the  meadow).  Seed  harvesting  be- 
gan somewhat  earlier  in  the  forest  than  in  the  meadow. 
In  fact,  about  50  percent  of  the  seed  crop  had  been  har- 
vested in  the  forest  by  August  31,  while  in  the  meadow 
this  point  was  not  attained  until  September  25.  As  seen 
on  the  graph,  no  seeds  remained  from  this  mast  year  in 
the  forest  stands  by  November  1.  Only  0.1  percent  of  the 
seeds  still  resided  in  the  tree  crowns  of  the  meadow  trees, 
and  these  were  no  longer  there  when  checked  on  27  June, 
1981. 

Unless  aided  by  an  animal,  cones  of  whitebark  pine  do 
not  fall  off  a  tree  to  the  ground  except  rarely  during  ex- 
tremely large  cone  crops — as  occurred  in  1989  (Hutchins 
and  Lanner  1982;  Lanner  1982;  Mattson  and  Reinhart, 
these  proceedings).  The  following  incident  shows  the  ex- 
tent Clark's  nutcracker  will  go  to  acquire  seed.  Hutchins 
and  Lanner  (1982)  placed  double-layer  wire  hardware 
screening  over  several  cones  so  there  would  be  cones  to 
collect  for  germination  studies.  By  late  October — after 
available  whitebark  pine  seeds  had  become  scarce — the 
nutcrackers  had  ripped  open  the  screening  and  harvested 
the  seed.  Seed  that  does  drop  to  the  ground  from  foraging 
accidents  made  up  only  4.2  percent  of  the  seed  crop  and 
most  of  this  seed  (69  percent)  was  determined  unviable 
through  examination  of  the  contents  in  the  shell  during 
1980  (Hutchins  1990). 

ANIMAL  INTERACTIONS  WITH  SEED 

A  variety  of  diurnal  animals  were  observed  active  in  or 
imder  whitebark  pine  trees.  Many  of  these,  however,  were 
never  observed  foraging  on  whitebark  pine  seed  in  this 
study.  These  include:  gray  jay  {Perisoreus  canadensis). 


common  flicker  (Colaptus  auratus),  Cassin's  finch  (Cora- 
podacus  cassinii),  rosy  finch  (Leucosticte  sp.),  pine  siskin 
(Spinas  pinus),  dark-eyed  junco  (Junco  hymelis),  black 
billed  magpie  (Pica  pica),  pine  marten  (Martes  americana), 
coyote  (Canis  latrans),  and  weasel  (Mustela  sp.). 

Many  feces  of  the  mammal  species  were  examined,  but 
no  evidence  of  whitebark  pine  seeds  was  found.  Gray  jays 
were  commonly  found  in  the  crowns  of  whitebark  pine  but 
were  only  observed  hawking  insects  and  caching  fresh 
carrion  or  boli  in  pine  branches.  Dow  (1965)  found  this 
species  has  little  interest  in  pine  seed  during  feeding  tri- 
als, although  Turcek  and  Kelso  (1968)  describe  its  Eur- 
asian cousin,  the  European  jay  (P.  infaustus),  as  having 
been  observed  taking  and  caching  Siberian  stone  pine 
(Pinus  sibirica). 

Of  the  birds,  only  the  magpie  could  potentially  be  in- 
volved in  the  dispersal  of  whitebark  pine  (Smith  and  Balda 
1979).  It  is,  however,  very  rare  in  the  subalpine  zone  dur- 
ing the  maturation  of  whitebark  pine  seed  (Hutchins  and 
Lanner  1982);  consequently,  any  potential  harvesting  of 
seed  would  be  unimportant  to  the  regeneration  of  white- 
bark pine. 

Many  species  have  been  observed  foraging  on  whitebark 
pine  cones  (Hutchins  1990,  table  2).  At  this  point  I  would 
like  to  take  a  closer  look  at  their  role  in  the  dispersal  and 
establishment  of  this  pine. 

BIRD  DISPERSERS 

Clark's  Nutcracker — Nutcrackers  were  the  most 
common  resident  bird  to  visit  the  whitebark  pine  trees 
(Hutchins  and  Lanner  1982).  They  were  scattered  about 
whitebark  pine  stands  in  loose  flocks  foraging  and  caching 
seed  of  this  tree.  They  are  dependent  on  these  caches 
year  round  (Giuntoli  and  Mewaldt  1978;  Vander  Wall 
and  Balda  1977;  Vander  Wall  and  Hutchins  1983). 


August 


September 


October 


November 


Figure  3 — Seasonal  course  of 
whitebark  pine  seed  harvest  by 
vertebrates  in  Squaw  Basin, 
WY,  in  1980. 


165 


Nutcrackers  were  observed  harvesting  seed  as  early 
as  July  13,  1979.  During  July,  however,  they  appeared  to 
only  be  testing  the  cones  for  ripeness  and  primarily  feed- 
ing on  the  previous  year's  caches  until  mid-August.  Seed 
harvested  at  this  time  were  lost  to  regeneration  because 
the  seed  coats  were  broken  and  the  seed  is  ungerminable 
vintil  August  13.  In  1980,  the  birds  were  able  to  success- 
fully harvest  whole,  developed  seed  by  August  15.  Nut- 
crackers harvested  seed  from  cones  at  increasingly  faster 
rates  (Hutchins  and  Lanner  1982)  through  early  October, 
when  seed  became  hard  to  find  (fig.  3).  Nutcrackers  were 
never  seen  attempting  to  harvest  Engelmann  spruce  or 
lodgepole  pine,  even  as  the  whitebark  pine  seed  crop  dwin- 
dled in  October.  As  the  seed  supply  became  scarce  in  mid- 
October,  I  observed  one  bird  checking  over  50  cones  for  a 
period  of  613  seconds  without  finding  a  single  seed. 

By  November  2,  it  appears  nutcrackers  were  almost 
totally  dependent  on  their  new  seed  caches,  and  would 
remain  so  imtil  the  following  August.  These  caches  were 
placed  just  below  the  soil  surface  (2-3  cm),  and  over  70  per- 
cent of  the  caches  were  either  one-  two-  or  three-seed 
caches  (x  =  3.2  +  2.8  seeds/cache;  fig.  4).  The  largest  cache 
was  14  seeds.  These  data  agree  closely  with  Tomback 
(1978)  and  Vander  Wall  and  Balda  (1977),  although  the 
number  of  single-seeded  caches  was  higher. 

Clark's  nutcrackers  cached  their  seeds  in  almost  every 
conceivable  soil  type  and  microsite  in  the  various  study 
sites.  Specific  examples  were  given  in  Hutchins  (1990). 
Although  they  may  cache  seeds  on  a  wide  variety  of  sites 
in  the  Rocky  Mountains,  several  studies  have  indicated 
that  south-facing  slopes  seem  to  be  preferred  (Lanner 
1982;  Lanner  and  Vander  Wall  1980;  Snethen  1980; 
Tomback  1978;  Vander  Wall  and  Balda  1977).  Seedling 
establishment  appears  to  be  much  more  common  on  more 
moist  sites  in  the  Rocky  Mountains  (Arno  1986;  Arno  and 
Hoff  1989;  Vander  Wall  and  Hutchins  1983)  than  in  the 
Sierra  Nevada  (Tomback  1982). 


40- 


0) 

o 

(0 

O 


c 

o 
o 

Q. 


35 
30 
25 
20 
15 
10 

5  ^ 


1    2    3    4    5    6    7     8    9   10   11  12   13  14 

Number  of  Seeds  per  Cache 

Figure  4 — Clark's  nutcracker  cache  size  fre- 
quency. Data  from  1 57  observations  made  at 
several  locations  in  Wyoming. 


Transport  distances  varied  greatly.  Seeds  were  placed 
as  close  as  50  m  from  the  site  of  the  harvested  tree,  or 
were  transported  at  least  3.5  km  to  the  Breccia  Cliffs  on 
the  edge  of  the  Squaw  Basin  site.  With  little  seed  left  in 
the  trees  by  mid-October,  the  birds  began  retrieving  many 
of  their  caches  made  in  the  Squaw  Basin  meadows  and 
recaching  them  on  the  southwest-facing  slopes  of  the 
Breccia  Cliffs. 

Large  flocks  of  nutcrackers  would  often  cache  seed 
together.  At  Mount  Washburn,  a  flock  estimated  at 
150  birds  was  seen  caching  seed  under  an  open-grown 
whitebark  pine  stand,  with  10-15  birds  within  a  10-m^ 
area.  No  aggression  occurred  among  the  birds  during 
these  observations. 

By  November  2,  1980,  snow  covered  many  of  the  cach- 
ing areas  for  the  winter.  Nutcrackers  were  observed  on 
several  occasions  successfully  pecking  through  up  to 
25  cm  of  snow  and  ice  to  retrieve  a  cache  during  the  win- 
ter months.  The  Togwotee  Pass  area  may  receive  up  to 
1,500  cm  of  snow  a  year,  yet  the  windswept  ridges  and 
south-facing  slopes  remain  exposed  enough  for  the  birds 
to  retrieve  their  caches.  Caches  on  the  northeast-facing 
slopes  and  under  the  forest  canopy  are  more  frequently 
used  during  June,  July,  and  August,  as  the  snow  recedes 
from  these  sites  last  (Vander  Wall  and  Hutchins  1983). 

The  number  of  seed  an  individual  nutcracker  caches 
annually  has  been  estimated  in  several  studies.  The 
numbers  for  whitebark  pine  range  from  32,000  in  the 
Sierras  (Tomback  1982)  to  98,000  in  the  Rocky  Moun- 
tains (Hutchins  and  Lanner  1982).  Because  of  the  num- 
ber of  variables  that  must  be  considered  (such  as  flight 
distances  and  the  amount  of  available  seed),  these  esti- 
mates vary  a  great  deal  from  site  to  site  and  year  to  year. 

Studies  by  Tomback  (1982),  Vander  Wall  (1988),  and 
Vander  Wall  and  Balda  (1977),  estimate  that  an  indi- 
vidual bird  caches  several  times  more  seed  than  it  needs 
to  survive  through  the  winter  and  early  spring.  At  that 
time  other  food  items  become  available — although  the 
nutcrackers  continue  to  use  their  caches  heavily  until 
the  new  cone  crop  begins  to  mature  (Vander  Wall  and 
Hutchins  1983).  This  leaves  many  unused  caches  for 
potential  germination  and  establishment  of  whitebark 
pine  seedlings  (Hutchins  and  Lanner  1982). 

Clark's  nutcracker  stands  out  from  all  the  other  poten- 
tial seed  dispersers  in  two  major  ways.  First,  it  consis- 
tently disperses  seed  in  a  way  that  increases  the  chance 
of  seedling  establishment.  The  seed  is  placed  just  below 
the  soil  surface  and  hidden  from  seed  predators  (Hutchins 
and  Lanner  1982;  Lanner  1980). 

Second,  nutcrackers  scatter  caches  across  the 
landscape — both  long  and  short  distances  from  the  source 
trees.  Dispersal  distances  of  up  to  22  km  have  been  ob- 
served by  Vander  Wall  and  Balda  (1977).  Also,  the  scat- 
tering of  thousands  of  their  caches  reduces  predation. 

Steller's  Jay  (Cyanocitta  stelleri) — Having  many  of 
the  behavioral  and  physiological  attributes  of  the  Clark's 
nutcracker  (Vander  Wall  and  Balda  1981),  the  Steller's 
jay  could  have  potential  as  a  dispersal  agent  of  whitebark 
pine.  In  this  part  of  their  range,  they  were  found  to  be 
primarily  solitary  foragers  and  uncommon  visitors  to  the 
whitebark  pine  forest.  These  birds  did  not  forage  for 
whitebark  pine  seed  until  early  September  when  the 


166 


cones  dried  and  the  scales  separated  slightly.  Their  bill 
structure  does  not  approach  the  efficiency  of  the  nut- 
cracker in  prying  apart  cone  scales  to  get  to  the  seed 
(Vander  Wall  and  Balda  1981),  thus  they  were  often  only 
able  to  extract  seed  from  cones  that  had  been  exposed  by 
nutcrackers.  These  jays  harvested  whitebark  pine  seed 
from  the  ground  and  canopy  in  less  than  half  of  their  ob- 
served foraging  time  during  the  fall  of  1980  and  1981 
(Hutchins  and  Lanner  1982).  The  seed  was  always  placed 
between  their  feet  and  hammered  with  their  bill  to  crack 
the  hull;  nutcrackers  primarily  cracked  the  seed  between 
their  mandibles. 

Occasionally  Steller's  jays  pouched  their  seed  in  their 
elastic  esophagus  for  later  caching.  They  were  observed 
caching  whitebark  pine  seeds  on  seven  occasions;  how- 
ever, none  of  these  were  in  the  soil.  The  birds  placed 
the  pine  seeds  in  the  crotch  of  a  tree,  a  densely  foliated 
witches  broom,  and  under  dense  lichen  growth  along  a 
tree  branch.  The  largest  number  of  seeds  I  observed  be- 
ing pouched  at  one  time  by  this  species  was  five,  although 
data  from  Vander  Wall  and  Balda  (1981)  indicate  these 
birds  could  hold  up  to  32  seeds/pouchload.  Because  this 
species  does  not  cache  whitebark  pine  seed  in  the  soil  and 
has  a  seed  foraging  rate  comparable  to  small  passerines 
(Hutchins  and  Lanner  1982),  it  is  a  very  improbable  agent 
for  seedling  establishment. 

Raven  (Corvus  corax) — third  corvid  rarely  observed 
foraging  on  whitebark  pine  seed  was  the  common  raven. 
These  birds  had  a  great  deal  of  difficulty  extracting  seed 
from  whitebark  pine  cones  due  to  their  thick,  long  bills. 
They  dropped  most  of  the  seed  when  foraging  (Hutchins 
and  Lanner  1982).  Caching  of  carrion  was  observed  with 
this  species,  but  no  observations  were  made  of  seed  cach- 
ing. In  Asia,  Reimers  (1959  cited  in  Tiircek  and  Kelso 
1968)  described  observations  of  ravens  caching  Japanese 
{Pinus  pumila)  and  Siberian  stone  pine  (P.  sibirica),  al- 
though it  was  not  clear  in  what  type  of  substrate. 

Other  Passerines — Pine  grosbeaks  (Pinicola  enuclea- 
tor)  primarily  foraged  on  whitebark  pine  seed  that  had 
been  exposed  by  Clark's  nutcrackers  breaking  off  cone 
scales.  Their  large  conical  beaks  also  enabled  them  to 
tear  away  cone  scales  to  get  at  seed.  Their  foraging  rates 
were  very  slow  compared  to  nutcrackers  due  to  their  tech- 
nique and,  as  a  consequence,  they  had  little  influence  on 
the  depletion  of  the  whitebark  pine  cone  crop. 

Grosbeaks  are  not  known  to  cache  food  (Smith  and 
Balda  1979;  Vander  Wall  1990).  During  my  observations 
of  this  species,  they  cracked  the  seed  coat  and  consumed 
the  seed  in  the  tree  crown.  It  is  highly  unlikely  that  they 
covild  pass  an  intact  seed  through  their  digestive  tract. 

The  moxmtain  chickadee  (Parus  gambeli)  and  red- 
breasted  nuthatch  (Sitta  canadensis)  occasionally  searched 
through  the  whitebark  pine  cones  during  September  after 
the  seeds  were  exposed  by  nutcrackers.  Both  these  spe- 
cies almost  always  dropped  the  seed  from  the  tree  crown, 
as  the  seeds  were  too  large  for  them  to  handle.  Their  un- 
successful foraging  bouts  contributed  to  the  small  percent- 
age of  seed  found  on  the  soil  surface,  which  was  later  con- 
sumed by  other  animals.  Neither  bird  species  was  observed 
caching  whitebark  pine  seed,  although  they  are  known  to 


cache  smaller  seeded  conifer  species  imder  bark  (Smith 
and  Balda  1979). 

MAMMAL  DISPERSERS 

Red  Squirrel  (Tamiasciurus  hudsonicus) — Red 

squirrels  are  common  denizens  of  whitebark  pine  forests. 
In  fact  they  were  the  second  most  commonly  observed  ver- 
tebrate next  to  the  Clark's  nutcracker  (Hutchins  and  Lan- 
ner 1982).  These  mammals  actively  defend  their  territo- 
ries (Smith  1968),  and  like  nutcrackers,  harvest  whitebark 
pine  cones  and  seeds  as  a  fall  and  winter  food  source.  Red 
squirrels  did  not  occiir  on  the  meadow  sites. 

Red  squirrels  spent  most  of  their  foraging  time  (75.8  per- 
cent) on  whitebark  pine  cones  or  recovering  dropped 
whitebark  pine  seeds.  This  includes  the  time  it  takes 
to  place  the  cones  in  the  midden.  Most  of  the  rest  of  the 
observed  foraging  activities  were  collecting  Engelmann 
spruce  cones  (11.3  percent)  and  harvesting  seeds  of  herba- 
ceous plants  (12.9  percent). 

Red  squirrels  began  harvesting  cones  as  early  as  nut- 
crackers harvested  seeds  (Jvily  13,  1979),  but  did  so  more 
intensely  during  July  while  the  nutcrackers  relied  more 
heavily  on  the  previous  year's  caches  (fig.  3).  Their  forag- 
ing rate  is  much  higher  than  that  of  the  other  mammals 
discussed  in  this  paper  because  they  usually  harvest  en- 
tire cones.  Most  of  the  time,  they  remove  the  cone  by  pull- 
ing it  off  the  branch  with  their  teeth,  instead  of  cutting 
the  subtending  branch  (Hutchins  1990).  Thus,  cone  re- 
moval is  likely  to  cause  little  change  in  tree  growth  form 
or  cone  production. 

Cone  caching  began  on  August  4  at  Squaw  Basin. 
Red  squirrels  often  would  leave  cones  on  the  groimd  below 
the  tree  where  they  were  cut  for  periods  of  up  to  3  days 
before  caching  them  in  their  midden.  Caching  of  other 
conifer  cone  species  began  later  with  Engelmann  spruce 
on  August  18  in  1980;  subalpine  fir  (September  11  in 
1980),  and  lodgepole  pine  (September  27  in  1980).  All 
cones  were  stored  in  "middens"  {n  =  114).  These  middens 
can  be  quite  extensive  and  are  composed  of  many  years  of 
cone  debris  above  the  soil  surface  (Finley  1969;  Reinhart 
and  Mattson  1990;  Smith  1970).  Of  the  time  spent  cach- 
ing food,  61.4  percent  was  devoted  to  whitebark  pine 
cones  and  16.9  percent  to  whitebark  pine  seeds.  Foraging 
on  the  cones  of  other  conifers  in  the  subalpine  made  up 
most  of  the  rest  of  the  caching  time  (7,304  seconds  of  ob- 
servation). Red  squirrels  also  cached  some  mushrooms 
and  herbaceous  seeds. 

Squirrels  were  also  found  to  cache  whitebark  pine  seeds 
in  their  midden  beginning  aro\md  September  16,  1980. 
These  seed  caches  were  placed  6.5-  to  40-cm  deep  {n  =  6); 
four  of  these  observations  were  between  11  and  11.5  cm 
deep.  The  number  of  seeds  per  cache  ranged  from  14  to 
55  ix  =  28.8;  19.2  seeds/cache;  n  =  4),  although  Kendall 
(1981)  foimd  up  to  176  seeds  in  a  single  hole.  I  examined 
the  seeds  from  two  squirrel  caches  gind  found  all  the  seeds 
sound. 

Red  squirrels  actively  chased  Clark's  nutcrackers  from 
the  trees  above  their  territories  and  their  middens.  On 
two  occasions,  however,  I  watched  nutcrackers  steal  whiter 
bark  pine  seeds  from  squirrel  middens. 


167 


Chipmunk  (Eutamias  sp.) — Chipmtinks  are  uncom- 
mon visitors  to  whitebark  pine  tree  crowns  but  do  occur 
on  both  open-meadow  and  forested  sites.  Most  of  their 
time  is  spent  on  the  ground,  usually  near  cover  plants  like 
sagebrush  (Artemisia  tridentata).  They  preferred  herba- 
ceous plant  parts  (lupine  seed,  grasses,  etc.)  on  my  subal- 
pine  study  sites  until  the  third  week  of  September  when 
these  plants  died  back  and  presimiably  lost  much  of  their 
nutritive  value.  At  that  time,  they  clumsily  foraged  on 
the  remaining  whitebark  cone  crop  (-10  percent  of  the 
seed  was  left  in  the  forest,  -40  percent  in  the  meadow). 
They  more  commonly  foraged  on  the  smsdl  amount  of  seed 
below  the  trees,  which  had  been  dropped  by  any  of  the 
species  discussed  earlier.  Tevis  (1953)  and  Heller  (1971) 
similarly  foimd  that  in  the  Sierra  Nevada  chipmxmks 
devote  little  time  to  foraging  on  whitebark  pine  seed  and 
cones. 

I  found  no  evidence  of  chipmimks  caching  whitebark 
pine  seed.  Broadbooks  (1958)  found  the  larder  cache 
depth  of  yellow  pine  chipmimks  {Eutamias  amoenus)  to 
average  28  cm.  They  use  these  types  of  caches  as  a  winter 
and  spring  food  source  when  they  are  periodically  aroused 
during  hibernation  (Vander  Wall  1990).  I  examined  chip- 
munk burrows  to  a  depth  of  20  cm  without  finding  any 
evidence  of  them  harvesting  and  caching  whitebark  pine 
seed. 

Yellow  pine  chipmunks  are  known  to  scatter-hoard 
smaller  amounts  of  Jeffrey  pine  (Pinus  jeffreyi)  and  pon- 
derosa  pine  {P.  ponderosa)  seed  in  shallower  caches  from 
1  to  25  mm  in  depth  (Vander  Wall  1992a).  These  species 
of  chipmunks  in  these  studies,  along  with  other  rodents, 
may  retrieve  many  caches  in  the  fall  in  the  Sierra  Nevada. 
They  use  the  seed  to  stock  their  winter  larders,  but 
many  seeds  survive  and  germinate  the  following  spring 
(Vander  Wall  1992b).  Although  chipmunks  may  play 
a  role  in  the  afforestation  of  certain  pines,  there  are  no 
data  that  support  this  conjecture  for  whitebark  pine. 
Vander  Wall  (1992a)  provides  strong  evidence  that  Jeffrey 
pine  may  arise  from  chipmunks  (Tamias  amoenus,  specio- 
sus,  and  quadramaculatus)  scatter-hoarding  seed.  In 
Japan,  Hayashida  (1989)  found  Siberian  chipmunks 
(T.  sihiricus)  were  not  an  effective  dispersal  agent  of 
Korean  stone  pine  {P.  koraiensis)  in  a  plantation.  In  this 
study,  the  limited  amount  of  seed  these  animals  harvest 
(see  seed  crop  depletion  section  of  this  paper)  apparently 
precludes  them  from  being  of  significance  to  whitebark 
pine  establishment  in  the  Rocky  Mountains.  More  detailed 
observations  of  chipmunks  in  whitebark  pine  ecosystems 
are  needed  to  more  accurately  define  their  role  as  cache 
predator  and  possible  seed  disperser. 

Golden-Mantled  Ground  Squirrel  (Spermophilus 
lateralis) — Due  to  a  slow  foraging  rate  and  relatively 
rare  occurrence  in  tree  canopies,  the  golden-mantled 
ground  squirrel  also  consumes  a  limited  amount  of  white- 
bark pine  seed.  This  species  was  found  in  a  whitebark 
pine  tree  only  once  in  188  hours  of  observation.  It  does 
collect  a  limited  amount  of  seed  from  the  ground,  which 
falls  from  cones  during  foraging  accidents  involving  other 
animals. 

Like  chipmunks,  groimd  squirrels  are  known  to  scatter- 
hoard  shallow  seed  caches  as  well  as  make  a  deeper  lar- 
der hoard  (MacClintock  1970;  Vander  Wall  1990).  This 


species  also  begins  hibernation  quite  early  in  September 
and  would  not  have  time  to  acquire  much  seed  for  storage. 
As  with  the  chipmimk,  few  studies  have  been  performed 
to  define  ground  squirrel  foraging  and  caching  behavior. 

SECONDARY  DISPERSERS 

Nocturnal  Rodents — This  group  of  animals,  made  up 
primarily  of  mice  and  voles  (family  Cricetidae),  must  also 
be  considered  as  potential  dispersal  agents  of  whitebark 
pine  seed.  They  were  not  directly  observed  in  this  study, 
but  possible  evidence  of  their  foraging  on  whitebark  pine 
seed  was  discovered  by  the  shelled  seed  left  behind  on  my 
simulated  cache  experiments. 

Smface  seed  caches  simulating  the  seed  that  was  foimd 
on  the  ground  indicate  it  will  not  last  long  (Hutchins  1989; 
McCaughey  and  Weaver  1990).  Most  of  the  seed  in  these 
caches  was  consumed  within  2  weeks  after  placement  un- 
der trees  by  evidence  of  the  seed  shells  left  behind.  Al- 
most all  of  the  shells  were  left  behind  at  the  cache  site, 
indicating  little  if  any  caching  was  done. 

There  are  two  primary  places  where  cricetid  rodents  may 
obtain  whitebark  pine  seeds:  (a)  from  seed  that  falls  to  the 
ground  and  (b)  the  discovery  of  animal  (primarily  nut- 
cracker) caches.  A  small  amount  of  seed  (-4  percent  of 
the  seed  crop)  was  found  to  fall  to  the  forest  floor  in  the 
mast  year  of  1980.  About  69  percent  of  seed  found  on  the 
ground  was  determined  imviable  by  examination  of  the 
contents  inside  the  shell.  Although  these  rodents  may 
recache  the  seed,  the  seed  was  probably  already  placed 
in  a  suitable  site  for  seedling  establishment  by  a  Clark's 
nutcracker.  Thus,  even  if  these  cricetids  relocate  the  seed 
to  another  cache,  their  positive  effect  on  the  establish- 
ment of  whitebark  pine  is  questionable  at  best. 

Abbott  and  Quink  (1970),  working  with  the  winged 
eastern  white  pine  {Pinus  strobus)  seeds,  showed  most 
caches  by  these  rodents  were  made  less  than  15  m  from 
the  seed  source.  Thus,  the  habit  of  whitebark  pine  trees 
pioneering  open  meadows  and  disturbed  areas  is  highly 
unlikely  to  arise  from  cricetid  caches.  Their  study  also 
showed  that  of  those  caches  not  recovered  by  the  time  the 
seed  germinated,  the  germinated  seedlings  were  soon  con- 
sumed by  these  animals.  This  information  coupled  with 
the  small  amount  of  seed  available  on  the  ground  for 
these  rodents  suggests  that  they  could  rarely  be  respon- 
sible for  seedling  establishment.  Future  studies  need  to 
look  at  this  group  of  dispersal  agents  more  closely  to  fur- 
ther delineate  their  role  in  whitebark  pine  ecosystems. 

Mice  of  the  genus  Peromyscus  are  members  of  this 
group  of  rodents,  which  is  known  to  scatter-hoard  pine 
seedhngs  (Abbott  and  Quink  1970;  Vander  Wall  1992a). 
Like  the  chipmunks  and  ground  squirrels,  this  species 
searches  for  seeds  on  the  ground  and  will  either  eat  them 
or  cache  them  for  later  use.  Because  so  few  sound  white- 
bark seeds  make  it  to  the  ground,  it  appears  the  role  of 
these  small  rodents  in  whitebark  pine  establishment  is 
insignificant. 

Grizzly  and  Black  Bears — Bears  {Ursus  arctos  and 
U.  americana)  primarily  obtain  seed  from  squirrel  middens 
(Kendall  1981,  1983;  Mattson  and  Jonkel  1989),  although 
black  bears  are  known  to  also  break  branches  to  harvest 
the  seed  (Tisch  1961).  Examinations  of  bear  scat  from 


168 


both  species  have  yielded  only  a  few  intact  whitebark  pine 
seeds  (Hutchins  and  Lanner  1982).  Even  if  these  seeds 
could  germinate  in  the  bear  scats,  they  would  produce  an 
insignificant  number  of  whitebark  pine  seedlings  that  are 
dispersed  a  very  short  distance  fi'om  the  squirrel  midden 
the  bear  raided. 

PERCENT  OF  SEED  HARVESTED 

On  forested  sites,  about  63  percent  of  the  seed  was  har- 
vested by  red  squirrels  and  36  percent  by  Clark's  nut- 
cracker. The  other  1  percent  was  harvested  by  all  other 
animals  combined  (table  1). 

In  contrast,  on  the  meadow  sites,  which  have  too  few 
trees  for  squirrels  to  exist,  nutcrackers  harvested  almost 
the  entire  crop  of  whitebark  pine  (99  percent).  Details  for 
how  these  values  were  calculated  are  described  in  Hutchins 
(1982,  1990). 

In  all  but  a  year  with  a  super-abundant  cone  crop,  ani- 
mals harvest  nearly  all  of  the  seed  crop  by  early  November 
in  the  Rocky  Moimtains  (Hutchins  and  Lanner  1982; 
Vander  Wall  1988).  By  this  time,  no  cones  remain  on 
the  trees  and  seeds  do  not  have  a  chance  to  survive  long 
enough  on  the  grotmd  £ind  germinate  as  suggested  by  Day 
(1967)  and  others.  In  this  study  I  found  that  the  seed  and 
cones  that  have  been  dropped  to  the  groimd  are  eaten 
within  3  weeks  by  various  foragers. 

CONCLUSIONS 

Nutcrackers,  squirrels,  and  the  other  whitebark  pine 
seed  forager  guilds  will  occasionally  drop  some  seed  to  the 
ground  (~4  percent).  Some  seed  is  dropped  intentionally 
(Vander  Wall  and  Balda  1977)  and  some  by  accident. 
Most  of  this  seed,  however,  is  composed  of  unfilled  or 
aborted  ovules  (69.5  percent)  and  could  never  germinate. 
Consequently,  dispersal  and  establishment  of  whitebark 
pine  by  small  rodents  like  deer  mice,  ground  squirrels, 
and  chipmunks  would  be  too  rare  an  event  for  a  tree  spe- 
cies to  survive  and  prosper. 

We  also  cannot  assert  red  squirrels  to  be  an  establisher 
of  whitebark  pine  seedlings.  As  I  stated  in  earlier  work, 
I  found  no  evidence  of  whitebark  pine  establishing  on 


squirrel  middens  (Hutchins  1989).  The  midden  "seed  bed" 
is  too  deep  in  organic  debris,  and  the  activity  on  the  midden 
constantly  distvu-bs  the  soil. 

Squirrels  do  harvest  most  of  the  whitebark  pine  seed 
in  the  forest,  and  they  are  the  major  seed  predator  on  for- 
ested sites.  Their  midden  cone  stores  are  also  important 
for  providing  grizzly  and  black  bears  with  an  important 
fall  food  source  (Reinhart  and  Mattson  1990). 

Nutcrackers,  on  the  other  hand,  are  one  of  the  most  im- 
portant biotic  influences  developing  and  changing  subal- 
pine  communities.  This  species  alone  probably  accoimts 
for  nearly  all  whitebark  pine  regeneration,  except  for 
chance  happenings.  As  far  as  the  bird  and  the  tree  are 
concerned,  it  is  more  profitable  to  cache  in  the  open  mead- 
ow. Much  less  predation  occurs  on  nutcracker  caches  in 
the  meadow  (Hutchins  1990),  and  the  small  ridges  are 
usually  fi'ee  of  snow  due  to  wind  action.  Consequently, 
the  higher  cache  survival  rate  benefits  tree  regeneration 
as  well  as  the  survival  of  the  nutcracker.  This  more  than 
any  factor  may  be  why  we  see  whitebark  growing  where 
we  do — pioneering  the  exposed  ridges,  roadside  cuts, 
burned  sites,  and  meadow  swales. 

The  open-grown  meadow  trees  may  be  the  most  impor- 
tant because  the  seeds  from  these  trees  will  have  a  greater 
chance  of  being  dispersed  by  Clark's  nutcracker.  These 
small  tree  islands  lack  the  major  seed  predator  (the  red 
squirrel).  It  should  be  added  that  Reinhart  and  Mattson 
(1990)  also  foimd  lower  numbers  of  red  squirrels  using 
pure  whitebark  pine  stands.  These  also  may  be  good 
sources  for  increasing  whitebark  pine  regeneration. 

When  nutcrackers  forget  where  they  placed  a  cache  (for 
example,  Vander  Wall  1982),  or  die,  or  a  rodent  does  not 
discover  the  seed  cache — it  has  a  chance  to  germinate. 
By  placing  the  seed  in  an  excellent  germination  bed  just 
below  the  soil  surface  (2-3  cm)  and  also  hiding  the  seed 
fi'om  easy  discovery  by  seed  predators,  the  bird  may  begin 
a  new  forest  stand. 

Eventually,  these  whitebark  pine  trees  modify  the  once- 
open  subalpine  landscape  so  other  more  shade-tolerant 
species  such  as  Engelmann  spruce  and  subalpine  fir  can 
establish  themselves  in  this  commimity  (Amo  1989; 
Franklin  and  Dyrness  1973;  Snethen  1980).  The  seed 
and  cover  produced  by  whitebark  pine  attracts  a  large 


Table  1 — Estimate  of  whitebark  pine  seed  harvested  by  various  animals  in  Wyoming 


Mean  seeds 

Minutes  spent 

Foraging 

Seeds 

Number  of 

Percent  seeds  harvested 

extracted/ 

foraging/ 

days/ 

harvested/ 

individuals 

by  all  individuals 

Species 

minute^ 

day^ 

season 

individual 

visiting  trees 

Forest 

Meadowr 

Clark's  nutcracker 

7.9 

180 

91 

129,402 

448 

36.3 

99.4 

Steller's  jay 

.7 

120 

55 

4,620 

11 

<.1 

.1 

Raven 

.6 

30 

53 

954 

15 

<.1 

<.1 

Noncorvids 

.7 

120 

56 

4,704 

43 

.1 

.3 

Red  squirrel 

43.4 

240 

84 

874,944 

116 

63.5 

Chipmunk 

1.7 

120 

35 

7,140 

10 

<.1 

.1 

'Seasonal  average  from  observations  made  during  August  15  to  October  11, 1980. 
^Estimate  made  from  observed  daily  activity  patterns. 
'Meadow  area  lacks  squirrels. 


169 


number  of  vertebrates  (Amo  and  Hoff  1989;  Hutchins 
1990;  Kendall  and  Arno  1990;  Lonner  and  Pac  1990; 
Tomback  1978). 

SUMMARY 

Whitebark  pine  is  dependent  on  animal  dispersal  for 
regeneration.  Many  animals  interact  with  whitebark  pine 
seed,  but  only  Clark's  nutcracker  was  found  to  consis- 
tently disperse  seed  in  a  way  that  leads  to  the  regenera- 
tion of  this  pine.  These  long-distance  dispersal  agents 
should  be  the  central  focus  in  whitebark  pine  subalpine 
community  management. 

ACKNOWLEDGMENTS 

I  thank  Steve  Vander  Wall  for  his  insight  on  small 
mammals  and  pine  seed  dispersal.  His  comments  were 
a  great  help.  Susan  Hutchins,  Ron  Lanner,  and  Diana 
Tomback  helped  with  earlier  versions  of  this  manuscript. 
I  thank  them  for  their  efforts. 

REFERENCES 

Abbott,  H.  G.;  Quink,  T.  F.  1970.  Ecology  of  eastern  white 
pine  seed  caches  made  by  small  mammals.  Ecology. 
51:  271-278. 

Arno,  S.  1986.  Whitebark  pine  cone  crops — a  diminishing 
source  of  wildlife  food?  Western  Journal  of  Applied  For- 
estry. 1(3):  92-94. 

Arno,  S.;  Hoff,  R.  J.  1989.  Pinus  albicaulis  Engelm.: 
whitebark  pine.  In:  Burns,  R.  M;  Honkala,  B.  H.,  tech. 
coords.  Silvics  of  North  America.  Vol.  1.  Conifers.  Agric. 
Handb.  654.  Washington,  DC:  U.S.  Department  of  Agri- 
culture, Forest  Service:  269-279 

Broadbooks,  H.  E.  1958.  Life  history  and  ecology  of  the 
chipmunk  {Eutamias  amoenus)  in  eastern  Washington. 
Misc.  Publ.  103.  Ann  Arbor,  MI:  Museum  of  Zoology, 
University  of  Michigan:  5-42. 

Day,  R.  J.  1967.  Whitebark  pine  in  the  Rocky  Mountains 
of  Alberta.  Forestry  Chronicle.  43:  278-282. 

Dow,  D.  D.  1965.  The  role  of  saliva  in  food  storage  by  the 
gray  jay.  Auk.  82:  139-154. 

Finley,  R.  B.,  Jr.  1969.  Cone  caches  and  middens  ofTam- 
iasciurus  in  the  Rocky  Mountain  region.  Misc.  Publ.  51. 
In:  Contributions  in  mammalogy.  Lawrence,  KS:  Uni- 
versity of  Kansas,  Museum  of  Natural  History:  233-274. 

Franklin,  J.  F.;  Dyrness,  C.  T.  1973.  Natural  vegetation 
of  Oregon  and  Washington.  Gen.  Tech.  Rep.  PNW-8. 
Portland,  OR:  U.S.  Department  of  Agriculture,  Forest 
Service,  Pacific  Northwest  Forest  and  Range  Experi- 
ment Station.  417  p. 

Giuntoli,  M.;  Mewaldt,  L.  R.  1978.  Stomach  contents 
of  Clark's  nutcracker  in  western  Montana.  Auk. 
95:  595-598. 

Hayashida,  M.  1989.  Seed  dispersal  by  red  squirrels  and 
subsequent  establishment  of  Korean  pine.  Forest  Ecol- 
ogy and  Management.  28:  115-129. 

Heller,  H.  C.  1971.  Altitudinal  zonation  of  chipmunks 
{Eutamias):  interspecific  aggression.  Ecology. 
52(2):  312-319. 


Hutchins,  H.  E.  1982.  The  role  of  birds  and  mammals 
in  the  dispersal  and  establishment  of  whitebark  pine. 
Logan,  UT:  Utah  State  University.  67  p.  Thesis. 

Hutchins,  H.  E.  1990.  Whitebark  pine  seed  dispersal  and 
establishment:  who's  responsible?  In:  Schmidt,  W.  C; 
McDonald,  K.  J.,  comps.  Proceedings — symposiimi  on 
whitebark  pine  ecosystems:  ecology  and  management  of 
a  high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  245-255. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Kendall,  K.  C.  1981.  Bear  use  of  pine  nuts.  Bozeman,  MT: 
Montana  State  University.  27  p.  Thesis. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  grizzly  and  black 
bears  in  the  Yellowstone  area.  Interagency  conference 
on  bear  research  and  management.  In:  Bears — their 
biology  and  management.  Calgary,  AB:  International 
Association  for  Bear  Management.  5:  166-173. 

Kendall,  K.  C;  Arno,  S.  F.  1990.  Whitebark  pine— an  im- 
portant but  endangered  wildlife  resource.  In:  Schmidt, 
W.  C;  McDonald,  K.  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  264-273. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  in  the  ecology 
and  evolution  of  limber  and  whitebark  pines.  In:  Sixth 
North  American  forest  biology  workshop  proceedings. 
Edmonton,  AB:  University  of  Alberta:  14-48. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for 
seed  dispersal  by  Clark's  nutcracker.  Canadian  Journal 
of  Forest  Research.  12(2):  391-402. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Joiirnal  of  For- 
estry. 78(10):  637-639. 

Lonner,  T.  N.;  Pac,  D.  F.  1990.  Elk  and  deer  use  of  white- 
bark pine  forests  in  southwestern  Montana:  an  ecologi- 
cal perspective.  In:  Schmidt,  W.  C;  McDonald,  K.  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
237-244. 

MacClintock,  D.  1970.  Squirrels  of  North  America.  New 
York:  Van  Nostrand  Reinhold.  184  p. 

McCaughey,  W.  W.;  Weaver,  T.  1990.  Biotic  and  micro- 
site  factors  affecting  whitebark  pine  establishment.  In: 
Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  Proceedings — 
symposivun  on  whitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource;  1989 
March  29-31;  Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest  Ser- 
vice, Intermountain  Research  Station:  140-150. 

Reinhart,  D.  P.;  Mattson,  D.  1990.  Red  squirrels  in  the 
whitebark  zone.  In:  Schmidt,  W.  C;  McDonald,  K.  J., 


170 


comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
256-263. 

Smith,  C.  C.  1968.  The  adaptive  nature  of  social  organiza- 
tion in  the  genus  of  tree  squirrels  Tamiasciurus.  Eco- 
logical Monographs.  38:  31-63. 

Smith,  C.  C.  1970.  The  coevolution  of  pine  squirrels 
{Tamiasciurus)  and  conifers.  Ecological  Monographs. 
40:  349-371. 

Smith,  C.  C;  Balda,  R.  P.  1979.  Competition  among  in- 
sects, birds,  and  mammals  for  conifer  seed.  American 

Zoologist.  19:  1065-1083. 
Snethen,  K.  L.  1980.  Whitebark  pine  {Pinus  alhicaulis) 

invasion  of  a  subalpine  meadow.  Logan,  UT:  Utah  State 

University.  76  p.  Thesis. 
Tevis,  L.,  Jr.  1953.  Stomach  contents  of  chipmunks  and 

mantled  ground  squirrels  in  northeastern  California. 

Journal  of  Mammalogy.  34:  316-324. 
Tisch,  E.  L.  1961.  Seasonal  food  habits  of  the  black  bear 

in  the  Whitefish  Range  of  Northwestern  Montana. 

Missoula,  MT:  University  of  Montana.  67  p.  Thesis. 
Tomback,  Diana  F.  1978.  Foraging  strategies  of  Clark's 

nutcracker.  The  Living  Bird.  16:  123-161. 
Tomback,  Diana  F.  1982.  Dispersal  of  whitebark  pine 

seeds  by  Clark's  nutcracker:  a  mutualism  hypothesis. 

Journal  of  Animal  Ecology.  51:  451-467. 


Turcek,  F.  J.;  Kelso,  L.  1968.  Ecological  aspects  of  food 
transportation  and  storage  in  the  Corvidae.  Commimi- 
cations  in  Behavioral  Biology.  1:  277-297. 

Vander  Wall,  S.  B.  1982.  An  experimental  analysis  of 
cache  recovery  in  Clark's  nutcracker.  Animal  Behaviour. 
30:  84-94. 

Vander  Wall,  S.  B.  1988.  Foraging  by  Clark's  nutcracker 
on  a  rapidly  changing  pine  seed  source.  Condor.  90: 
621-631. 

Vander  Wall,  S.  B.  1990.  Food  hoarding  in  animals. 
Chicago,  IL:  University  of  Chicago  Press.  445  p. 

Vander  Wall,  S.  B.  1992a.  Establishment  of  Jeffrey  pine 
seedings  from  animal  caches.  Western  Jovirnal  of  Ap- 
plied Forestry.  7(1):  14-20. 

Vander  Wall,  S.  B.  1992b.  [Personal  communication]. 
July.  Reno,  NV:  University  of  Nevada,  Department 
of  Biology. 

Vander  Wall,  Stephen  B.;  Balda,  R.  P.  1977.  Coadapta- 
tions  of  Clark's  nutcracker  and  the  pinon  pine  for  effi- 
cient seed  harvest  and  dispersal.  Ecological  Monographs. 
47:  89-111. 

Vander  Wall,  Stephen  B.;  Balda  R.  P.  1981.  Ecology 
and  evolution  of  food-storage  behavior  in  conifer-seed 
caching  corvids.  Zeitschrift  fiir  Tierpsychologie.  56: 
217-242. 

Vander  Wall,  Stephen  B.;  Hutchins,  Harry  E.  1983. 
Dependence  of  Clark's  nutcracker,  Nucifraga  Colum- 
biana, on  conifer  seeds  during  the  postfledging  period. 
The  Canadian  Field  Naturahst.  97(2):  208-214. 


171 


SIZE  OF  PINE  AREAS  IN  RELATION 
TO  SEED  DISPERSAL 


Hermann  Mattes 


Abstract — Pine  seeds  are  either  wind-  or  bird-  (animal-) 
dispersed.  Only  well-adapted  seeds  support  large  ranges,  which 
are  of  similar  size  in  both  types  of  dispersal.  Seeds  adapted  to 
wind  dispersal  should  have  a  ratio  of  seed  weight  to  wing  length 
below  1.8  mg/mm.  Bird-dispersed  seeds  have  a  weight  between 
90  and  550  mg.  Species  with  seeds  above  100  mg  but  without 
zoochorous  features  have  small  or  even  relictic  ranges.  Large 
seeds  are  advantageous  under  severe  conditions  and  in  competi- 
tion. Consequently,  the  bird-pine  mutualism  contributes  to 
larger  ranges  of  those  pine  species. 


Distribution  ranges  of  plant  and  animal  species  are  de- 
termined by  four  main  factors:  ecological  potency  of  the 
species;  effectiveness  of  seed  dispersal;  size  of  area  avail- 
able; and  time  available.  Ecological  studies  on  plants  and 
vegetation  mostly  focus  on  site  conditions.  Indeed,  compe- 
tition for  nutrients,  water,  light,  and  so  on  most  often 
turn  out  to  be  the  main  factor  limiting  distribution.  How- 
ever, first  of  all,  seeds  must  be  able  to  reach  the  site  in 
question.  Therefore,  seed  dispersal  is  an  important  factor 
in  plant  ecology.  Patchy  and  irregular  patterns  of  species 
on  ruderal  or  fallow  ground  are  a  result  of  dissemination. 
Otherwise  vegetation  history  in  the  Holocene  gives  evi- 
dence that  time  since  last  glaciation  was  not  long  enough 
for  full  expansion  of  all  species.  We  expect  effectiveness 
of  seed  dispersal  to  be  correlated  with  rapidness  of  expan- 
sion and  with  size  of  range. 

Small,  winged  seeds  are  easily  disseminated  by  wind, 
which  is  present  almost  everywhere  and  emytime.  Com- 
petition of  the  seedling  within  a  closed  vegetation  cover 
as  well  as  harsh  environmental  conditions  require  large 
seeds  with  a  high  amount  of  nutrient  reserves.  Many 
pine  species  have  developed  an  almost  obligate  mutualis- 
tic  relationship  to  birds  (Nucifraga,  Gymnorhinus)  for 
seed  dispersal.  Pines  with  bird-dispersed,  large  seeds  be- 
came dominating  tree  species  at  arctic  and  alpine  timber 
lines,  in  xeric  environments  as  well  as  in  mesic  lowland 
forests. 

The  genus  Pinus  with  more  than  100  species  provides 
a  good  chance  to  test  effectiveness  of  seed  dispersal.  All 
pine  seeds  are  of  a  rather  similar  structure;  the  most  obvi- 
ous differences  are  seed  size  and  wing  length.  The  pri- 
mary type  is  very  likely  an  anemochorous  seed  with  a 
long  wing.  Second,  many  species  have  developed  large 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Hermann  Mattes  is  Professor,  Institut  fur  Geographie,  Abteilung 
Landschaftsokologie,  Westfalische  Wilhelms-Universitat,  Robert-Koch- 
Str.  26,  W-4400  MOnster,  Germany. 


seeds  with  shortened  or  missing  vnngs.  These  seeds  are 
disseminated  mainly  by  birds,  or  occasionally  by  squirrels. 
Apart  from  that,  seeds  and  cones  show  many  morphologi- 
cal and  phenological  adaptations  for  dispersal,  which  are 
discussed  elsewhere  (Lanner  1980,  1982;  Vander  Wsdl  and 
Balda  1977). 

In  this  paper  some  relations  of  seed  characteristics  and 
size  of  ranges  are  discussed.  The  main  hjrpothesis  is  to 
find  the  larger  ranges  within  pine  species  well  adapted 
to  seed  dispersal. 

MATERIALS  AND  METHODS 

From  65  out  of  about  106  pine  species  in  total  enough 
detailed  information  is  available  for  a  comparison  of  dis- 
persal abilities  and  range  sizes.  The  pine  species  are 
numbered  following  Mirov  (1967). 

Distribution  ranges  are  of  a  complex  nature.  Effective- 
ness of  dispersal  would  be  measured  best  by  the  propor- 
tion of  the  potential  area  that  really  has  been  occupied  by 
the  species.  Unfortunately,  we  do  not  know  exactly  the 
potential  ranges  of  almost  all  plant  species.  An  interest- 
ing idea  would  be  to  measure  the  distances  or  Eireas  occu- 
pied after  the  end  of  the  last  glaciation.  However,  refugials 
and  paths  of  dispersal  are  very  incompletely  known. 
What  we  can  clearly  see  are  area  and  distances  within 
ranges.  However,  it  seems  not  to  be  adequate  to  use  the 
area  of  a  range.  It  is  highly  influenced  by  size  of  the  re- 
gion (the  continent  or  vegetation  zone  that  is  inhabited 
by  the  species  concerned).  Error  should  be  less  using  the 
distance  between  the  outermost  points  of  a  range.  This 
was  calculated  as  the  orthodrome  distance. 

Split  or  disjunct  ranges  have  been  treated  in  the  same 
way  as  continuous  ranges.  It  has  proved  to  be  too  difficult 
to  decide  whether  an  interrupted  range  was  due  to  natu- 
ral factors  or  to  hxmian  influences,  especially  in  regions 
that  were  densely  populated  for  a  long  time  such  as  East 
Asia  and  the  Mediterranean. 

In  the  present  study  all  105  pine  species  listed  in  Mirov 
(1967),  and  in  addition  Pinas  longaeva,  are  considered. 
Subspecies  have  not  been  considered.  Reinge  maps  £ind 
information  was  taken  from  Kriissmann  (1968),  Little 
(1971),  Meusel  and  others  (1965),  and  Mirov  (1967). 

In  only  a  few  experiments  flying  ability  of  pine  seeds 
was  examined  (Lanner  1985;  Muller-Schneider  1977). 
To  estimate  dispersal  abilities  of  numerous  species,  indi- 
rect methods  have  been  used.  Simple,  but  reliable  meas- 
lu-ements  for  anemochorous  seeds  are  weight  of  cleaned 
seeds  and  wing  length.  Such  data  are  available  for  most 
of  the  species  (Kriissmann  1972;  USDA  1974).  The  ratio 
of  seed  weight  to  wing  length  is  used  as  an  indicator  for 
potency  of  dispersal.  Volume  and  surface  of  the  seed  and 
shape  and  surface  of  the  vring  would  be  of  great  interest, 


172 


but  these  data  are  not  available  for  most  species.  Zooch- 
orous  species  are  even  more  difficult  to  evaluate.  Seed 
and  cone  characteristics  of  species  fully  developed  for  bird 
dispersal  are  compared  morphologically  with  less-adapted 
species. 

DISPERSAL  OF  PINE  SEEDS 

Seed  dispersal  either  by  wind  or  by  birds  has  caused 
specific  adaptions.  Most  of  the  pine  species  can  be  attrib- 
uted to  these  two  main  types.  A  narrow  seed-weight  to 
wing-length  ratio  is  considered  to  be  favorable  to  flight 
ability.  From  figure  1  with  46  pine  species  Usted  we  can 
assume: 

•  In  general,  the  length  of  seed  wings  increases  until 
seed  weight  reaches  about  90  mg  (P.  palustris:  93  mg/ 
36  mm);  however,  the  ratio  of  seed  weight  to  wing  length 
increases  obviously  with  increasing  seed  weight. 

•  Winged  seeds  of  more  than  about  100  mg  have  an 
unfavorably  high  ratio  of  seed  weight  to  wing  length;  wing 
length  is  insvifficient  even  in  species  with  huge  cones  and 
long  cone  scales.  These  seeds  are  expected  to  have  no 
adequate  flight  abilities. 

•  The  smallest  seed  weight  of  a  wingless  seed  species 
{P.  flexilis)  is  93  mg;  most  of  the  species  with  heavier 
seeds  are  wingless,  or  seed  wings  separate  easily  (P.  penta- 
phylla  No.  83,  P.  himekomatsu  No.  84,  P.  strobiformis 
No.  35;  also  in  P.  pinea  No.  73). 

Comparing  morphological  features  with  area  size  we 
can  conclude  the  following  (fig.  2,  table  1). 

•  Wind-dispersed  pine  species  with  a  low  ratio  of  seed 
weight  to  wing  length  often  have  larger  distribution 
ranges  than  species  with  a  high  ratio. 

•  There  is  evidence  for  a  log-linear  relationship  for  the 
largest  ranges  of  each  class  of  seed  weight/wing  length  ra- 
tio. Species  with  a  ratio  higher  than  2  mg/mm  are  less 
distributed  on  an  average.  It  is  supposed  that  dispersal 
abilities  limit  the  range  of  distribution.  A  remarkable  ex- 
ception is  P.  pinea  (No.  73).  Its  origin  is  not  known,  and 
it  has  been  cultivated  in  many  places  since  Phoenicean 
times,  approximately  4,000  years  ago.  Therefore,  the 
whole  present  range  had  to  be  taken  into  consideration. 

•  Many  species  do  not  reach  an  area  as  large  as  could 
be  expected  in  view  of  their  theoretical  abilities  for  seed 
dispersal.  This  can  be  partly  explained  by  geographical 
and  ecological  reasons.  Species  of  southeastern  North 
America  (Nos.  22-31)  depend  on  a  relatively  small  area 
limited  by  the  ocean  or  woodless  plains.  The  range  of 
Cahfomian  pines  (Nos.  5,  6,  17-19)  of  the  Mediterranean 
type  is  limited  by  climate.  Mexican  pines  (seed  data  only 
for  P.  chihuahua  and  patula,  Nos.  40  and  61)  also  are  re- 
stricted geographically  and  ecologically  to  narrow  moxm- 
tainous  belts.  Pinus  peuce  and  P.  heldreichii  (Nos.  71  and 
75)  are  endemic  to  the  Balkan  mountains  in  southeastern 
Europe. 

•  Ranges  of  bird-dispersed  pines  are  of  similar  size  as 
those  of  wind-dispersed  pines;  differences  are  mainly  due 
to  the  very  restricted  range  of  the  bird-dispersed  P.  quad- 
rifolia  in  America  and  the  huge  range  of  wind-dispersed 
P.  sylvestris  in  Eurasia.  Some  bird-dispersed  pine  ranges 
are  among  the  largest  of  all  pines.  Concerning  size  of 


ranges,  dispersal  by  birds  (and  sqmrrels)  is  as  efficient 
as  dispersal  by  wind. 

•  Pine  ranges  in  America  are  smaller  than  those  in 
Eurasia.  This  is  an  effect  of  the  larger  land  mass  in  the 
Old  World  hemisphere.  Only  species  with  well-adapted 
dispersal  mechanisms  are  affected;  species  with  relicted 
ranges  show  no  difference  in  area  size. 

GENUS  PINUS  DISTRIBUTION 

Most  of  the  pine  species  are  holarctic.  Pines  occupy 
a  wide  variety  of  sites  of  which  the  most  extreme  are 
located  at  timber  lines  in  cold  and  arid  climates  as  well 
as  in  very  dry  or  perhvunid  regions  in  the  subtropics  and 
tropics.  Many  of  the  pine  species  are  restricted  to  specific 
sites.  Thus,  an  effective  dispersal  agent  is  needed  to 
reach  suitable  sites. 

Recent  centers  of  species  diversity  (fig.  3)  are  Mexico 
and  the  southwestern  States  of  the  U.S.A.  Secondary  cen- 
ters of  diversity  are  eastern  North  America,  eastern  Asia, 
and  the  Mediterranean  region.  These  were  refugial  areas 
during  glaciation.  Although  none  of  these  areas  is  neces- 
sarily the  origin  place  of  the  genus  Pinus,  recent  regions 
of  high  species  numbers  of  pines  are  assumed  to  be  places 
where  pines  have  existed  for  a  long  time. 

In  contrast,  there  are  large  areas  populated  only  by  a 
few  pine  species.  This  is  the  case  especially  in  the  present 
boreal  zone.  Also,  we  find  relatively  few  pine  species  in 
Central  America  south  of  Tehuantepec,  on  the  Caribbean 
Islands,  in  the  Himalayas,  and  in  southeastern  Asia. 
Probably,  these  areas  have  been  occupied  by  pines  for 
only  a  relatively  short  time. 

The  boreal  zone  was  reoccupied  by  pines  in  the  postgla- 
cial time,  and  we  know  that  some  species  are  stiU  expand- 
ing (for  example  see  Gorchakovsky  1993,  for  Pinus  sibi- 
rica).  Nevertheless,  their  ranges  belong  to  the  largest 
ranges  of  pine.  All  boreal  pine  species  are  among  the  spe- 
cies best  adapted  either  for  wind  dispersal  (figs.  1  and  2; 
Nos.  16,  20,  21,  32,  and  69)  or  bird  dispersal  (Nos.  67  and 
68).  Wind-dispersed  pines  have  excellent  seed-weight/ 
wing-length  ratios  from  0.4  to  1.1  mg/mm. 

Data  are  available  for  only  a  few  species  of  southern 
marginal  regions  (P.  wallichiana,  No.  91;  P.  merkusii, 
No.  101;  P.  khasya,  No.  104).  With  a  ratio  of  seed  weight/ 
wing  length  of  1.2  to  1.7  mg/mm,  adaptation  for  wind  dis- 
persal in  these  species  is  relatively  good. 

Species  that  were  more  widely  distributed  in  the  Ter- 
tiary or  early  Pleistocene,  and  that  now  occupy  relictic 
ranges  only,  show  very  different  ratios  of  seed  weight/ 
wing  length.  For  P.  aristata  (No.  5;  1.2  mg/mm)  and  per- 
haps P.  balfouriana  (No.  6;  2.1  mg/mm)  and  P.  longaeva 
(No.  6a)  seed  dispersal  might  not  be  a  restricting  factor. 
This  svirely  is  the  case  in  P.  torreyana  (No.  11). 

Species  with  reduced  seed  wings  have  less  chances  to 
be  dispersed  by  wind.  Some  of  them  are  on  the  way  to 
developing  zoochorous  features.  An  advanced  state  of 
zoochory  shows  in  especially  P.  strobiformis  (No.  35),  and 
somewhat  less  in  P.  pentaphylla  (No.  83)  and  P.  himeko- 
matsu (No.  84)  (Hayashida  1989);  they  have  large  seeds 
and  wings  that  easily  separate  fi-om  the  seed.  These  spe- 
cies are  already  mainly  disseminated  by  birds  and  ani- 
mals. Pinus  peuce  (No.  71)  has  a  small  seed  (40  mg)  and 


173 


length  of  seed  wing 


0.4 
0.4 

0.5  n 

0.4  ] 
0.4  D 
0.7 
0.6 
0.S 
0.6  h 
0.7] 
1.6  3 

1.13 
1.1  J 


0.7  1 
1.5  □ 


16 
5.9 

n  84 


seed  weight 

banks iana  32 
contorta  16 
sylvestris  69 
clausa  28 
serotina  30 
pa tula  61 
resinosa  21 
virginiana  27 
chihuahua  40 
pungens  31 
halepensis  77 
strobus  20 
nigra  74 

1.0  p  monticola  4 
attenuata  17 
heldreichii  75 

1.3  p  merkusii  101 

2.1  □  aristata  6 

1.3  ZJ  elliottii  23 
1.5  tzi  radiata  19 

ponderosa  14 
peuce  71 
engelmannii 
pinaster  80 
wall ichi ana 

□  palustris 

□  flexilis  2 

□  pumila  6  8 
m!  pentaphylla  83 
~i  jeffreyi  13 

:l  strobiformis  35 
I  albicaulis  1 
1  lamber tiana  3 


1.3 
5.7 


1.6  i 

1.9  ZD 


1.7 
2.6 


7.7 


13 


[j378 
56 


53 

91 
22 


181 


cembra  7  0 
edulis  8 
j  sibixica  67 

U  armandii  86 
i   coulteri  12 

.  1  monophylla  7 

1  gerardiana  92 
"    I  cembroides  36 
•  ^   I  quadrifolia  9 

:    "~1  koraiensis  82 

I  pinea  7  3 
:  .  '  I  sabiniana 

torievana] ll 


10 


[mml  30  20 


10 


0     0      100    200    300     400      500    600    700     800      900  [mgl 


Figure  1— Wing  length  (mm,  left),  weight  (mg,  right),  and  ratio  of  wing  length  to  weight  (mm/mg, 
middle)  of  seeds  of  46  pine  species.  Numbers  following  species  names  according  to  Mirov  (1967). 
For  sources  see  text.  Species  are  arranged  according  to  seed  weight. 


is  not  very  attractive  for  nutcrackers,  although  it  is  used 
for  hoarded  food  sometimes  as  is  P.  longaeva  (Lanner 
1988).  It  is  assumed  that  dispersal  by  nutcrackers  is  of 
some  relevance  in  establishing  these  two  species  at  tim- 
berline.  Ranges  of  both  species  are  small. 


Other  pine  species  such  as  P.  lambertiana,  P.  sabiniana, 
P.  torreyana,  P.  coulteri,  and  P.  jeffreyi  (Nos.  3,  and  10-13) 
have  large  seeds  but  lack  further  adaptations  for  disper- 
sal by  birds  and  animals.  Since  these  species  have  no  ef- 
fective dispersing  agent,  their  ranges  are  relatively  small. 


174 


{mg/mmj  in 
403  6-j 


^  148 
"5 

55 
5  7.4 

f 

2.7 
1.0 


5  -1 
4 
3  - 
2  H 

1 
0 


0.4  -1  -i 
0.6  -1.5 


11 


71 


•63 


7      36  1  86 
ftX'n   **    ****     •    ■  * 

92  8    70  35  2     8  2    6  7  6  8 


'  73 


84 


•  83 


'12 


•  3 

•  13 


•  Ca 


.  53 


-91 


•  78 
•22 

5     '19,  24';i-0l'77 
23  1  a    .  •    •  •14 

40.,7.95  'ZS^^ 

.28  26  ^52^  .32 
•l6 


•69 


5.5 
246 


6.0 
404 


6.5 
685 


7.0     7.5      8.0      8.5      9.0  9.5 


In 


1097  1808   2961    4815    8103  13360  [km] 
size  of  range 


Figure  2 — Size  of  ranges  in  relation  to  the  ratio  wing  length/ 
seed  weight  of  44  wind-dispersed  pines;  for  comparison, 
range  size  of  13  bird-dispersed  pine  species  on  top  of  the 
figure.  For  code  numbers  for  pine  species  see  appendix; 
for  sources  see  text. 


PiniLs  pinea,  as  mentioned  earlier,  is  an  exception  because 
it  has  been  cultivated  throughout  history. 

CONCLUSIONS 

Competition  in  seedlings  causes  a  high  selective  pres- 
sxire  for  large  seeds.  Even  among  wind-dispersed  pines 
seed  size  tends  to  enlarge  despite  a  strong  negative  effect 
on  dispersal  ability.  Obviously,  at  90  to  100  mg  seed 
weight  dispersal  success  by  wind  has  diminished  so  far 
that  other  agents  are  essential.  From  24  haploxylon  pine 
species,  19  species  have  wingless  or  almost  wingless  and 


large  seeds,  which  are  proven  to  be  or  expected  to  be  bird 
dispersed.  There  is  no  doubt  that  zoochory  has  developed 
in  parallel  in  several  sections  of  haploxylon  pines  (Lanner 
1989;  Tomback  and  Linhart  1990).  Three  diploxylon  spe- 
cies (P.  torreyana,  No.  11;  P.  sabiniana,  No.  12;  P.  pinea, 
No.  73)  have  large,  nearly  wingless  seeds  but  lack  further 
adaptations  for  zoochory. 

Range  sizes  of  bird-dispersed  pines  are  not  obviously 
different  from  those  of  wind-dispersed  pines.  Dissemina- 
tion of  pine  seeds  by  birds  is  at  least  as  effective  as  wind 
dispersal.  The  interaction  of  pines  and  nutcrackers  or 
jays  is  well  balanced.  Even  scattered  stands  of  zoochorous 


Table  1 — Average  size  of  ranges  (km  distance)  of  pine  species  with  different  dispersal  agents.  Wind-dispersed 
pines  are  arranged  according  to  their  seed  characteristics  (seed  weight/wing  length  in  mg/mm) 


 Distribution  ranges  

Pine  species  dispersed  by:  America  Eurasia  Global 


Birds 

1,557 

in 

=  7) 

2,680 

in  = 

6) 

2,075 

in 

=  13) 

Wind  (0.4-0.8  mg/mm) 

1,913 

in 

=  13) 

M,247 

in  = 

3) 

2,351 

in 

=  16) 

Wind  (0.9-2.0  mg/mm) 

1,926 

in 

=  9) 

2,610 

in  = 

7) 

2,225 

in 

=  16) 

Wind  (>2.0  mg/mm) 

1,121 

in 

=  7) 

21,055 

in  = 

4) 

21,097 

in 

=  11) 

Species  considered  above 

1,693 

in 

=  36) 

22,566 

in  = 

20) 

21,994 

in 

=  57) 

All  species 

1,413 

in 

=  67) 

1,779 

in= 

39) 

1,548 

in 

=  106) 

'Without  P.  sylvestris,  average  range  size  is  2,150  km. 
^Numbers  without  P.  pinea. 


175 


176 


pine  trees  support  all  needs  for  territories  of  nutcrackers 
concerning  breeding,  feeding,  overwintering,  and  social 
life.  Nutcrackers  and  jays  have  still  preserved  character- 
istic euryoecious  features  of  corvids,  especially  omnivory. 
In  Eurasia,  where  there  is  almost  no  other  animal  com- 
peting with  the  nutcracker  for  seed  hoarding,  the  range 
of  the  nutcracker  surpasses  the  ranges  of  zoochorous 
pines  in  many  regions. 

Within  wind-dispersed  pines  we  find  that  large-sized 
ranges  suppose  small  seeds  with  a  ratio  of  weight  to  wing 
length  below  approximately  1.8  mg/mm,  but  not  vice- 
versa  because  of  geographical  barriers  or  specialized  fea- 
tures of  a  species.  However,  pine  species  with  a  very  high 
ratios  of  seed  weight/wing  length  (above  1.8  mg/mm)  al- 
ways have  restricted  areas. 

Altogether,  pine  seed  dispersal  by  birds  was  very  suc- 
cessful dvuing  evolution.  Pine  species  became  able  to  oc- 
cupy additional  and  more  different  sites,  as  well  as  under 
extreme  conditions  at  the  borders  of  tree  growth  as  under 
mesic  conditions.  Consequently,  bird-dispersed  pines 
were  able  to  enlarge  their  ranges. 

REFERENCES 

Critchfield,  William  B.;  Little,  Elbert  L.,  Jr.  1966.  Geo- 
graphic distribution  of  the  pines  of  the  world.  Misc. 
Pub.  991.  Washington,  DC:  U.S.  Department  of  Agricul- 
ture, Forest  Service.  97  p. 

Gorchakovsky,  P.  L.  1993.  Distribution  and  ecology  of  Si- 
berian stone  pine  (Pinus  sibirica)  in  the  Urals.  (These 
proceedings). 

Hayashida,  Mitsuhiro.  1989.  Seed  dispersal  and  regenera- 
tion patterns  of  Pinus  parviflora  var.  pentaphylla  on 
Mt.  Apoi  in  Hokkaido.  Research  Bulletin  of  the  College 
Experiment  Forests,  Faculty  of  Agriculture,  Hokkaido 
University.  46(1):  177-190. 

Kriissmann,  Gerd.  1968.  Die  Baume  Europas.  Berlin  and 
Hamburg:  Parey  Verlag.  140  p. 

Kriissmann,  Gerd.  1972.  Handbuch  der  Nadelgeholze. 
Hambiu-g  und  Berlin:  Parey- Verlag. 

Lanner,  Ronald  M.  1980.  Avian  seed  dispersal  as  a  factor 
in  the  ecology  and  evolution  of  limber  and  whitebark 


pines.  Proc.  6th  North  American  Forest  Biology  Work- 
shop, University  of  Alberta,  Edmonton,  Canada:  15-48. 

Lanner,  Ronald  M.  1982.  Adaptations  of  whitebark  pine 
for  seed  dispersal  by  Clark's  nutcracker.  Canadian 
Journal  of  Forest  Research.  12(2):  391-402. 

Lanner,  Ronald  M.  1985.  Effectiveness  of  the  seed  wing  of 
Pinus  flexilis  in  wind  dispersal.  Great  Basin  Naturalist. 
45(2):  318-320. 

Lanner,  Ronald.  1988.  Dependence  of  Great  Basin  bristle- 
cone  pine  on  Clark's  nutcracker  for  regeneration  at  high 
elevation.  Arctic  and  Alpine  Research.  20(3):  358-362. 

Lanner,  Ronald  M.  1989.  Biology,  taxonomy,  evolution, 
and  geography  of  stone  pines  of  the  world.  In:  Schmidt, 
Wyman  C;  McDonald,  Kathy  J.,  comps.  Proceedings — 
symposium  on  vv^hitebark  pine  ecosystems:  ecology  and 
management  of  a  high-mountain  resource.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
14-24. 

Little,  Elbert  L.,  Jr.  1971.  Atlas  of  United  States  trees. 

Vol.  1:  Conifers  and  important  hardwoods.  Misc.  Publ. 

1146.  Washington,  DC:  U.S.  Department  of  Agriculture. 

9  p.  and  200  maps. 
Meusel,  H.;  Weinert,  E.;  Jager,  E.  1965.  Vergleichende 

Chorologie  der  zentraleuropaischen  Flora.  Bd.  I.  Jena: 

Fischer- Verlag.  9  p.  and  200  maps. 
Mirov,  Nicholas  T.  1967.  The  genus  Pinus.  New  York: 

Ronald  Press.  602  p. 
Miiller-Schneider,  Paul.  1977.  Verbreitungsbiologie 

(Diasporologie)  der  Blutenpflanzen.  Veroffentlichungen 

des  geobotanischen  Institutes  der  ETH  Zurich,  Stiftung 

Riibel.  Vol.  61.  226  p. 
Tomback,  Diana  F.;  Linhart,  Yan  B.  1990.  The  evolution 

of  bird-dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 
U.S.  Department  of  Agriculture,  Forest  Service.  1974. 

Seeds  of  woody  plants  in  the  United  States.  Agric. 

Handb.  450.  Washington,  DC.  838  p. 
Vander  Wall,  Stephen  B.;  Balda,  Russ  P.  1977.  Coadapta- 

tions  of  the  Clark's  nutcracker  and  the  pinyon  pine  for 

efficient  harvest  and  dispersal.  Ecological  Monographs. 

Durham.  47(1):  27-37. 


177 


APPENDIX:  CODING  LIST  OF  NUMBERS  OF  PINE  SPECIES  CONSIDERED 
(AFTER  MIROV  1967)  wi^c»ii^r.rtr.u 


Western  North  America 

Haploxylon  pines 

1  albicaulis 

2  flexilis 

3  lambertiana 

4  monticola 

5  balfouriana 

6  aristata 
6a  longaeva 

7  monopylla 

8  edulis 

9  quadrifolia 

Diploxylon  pines 

10  sabiniana 

11  torreyana 

12  coulteri 

13  jeffreyi 

14  ponderosa 

16  contorta 

17  attenuata 

18  muricata 

19  radiata 

Eastern  North  America 

Haploxylon  pines 

20  strobus 

Diploxylon  pines 

21  resinosa 

22  palustris 

23  elliotti 

24  toeda 

25  echinata 

26  glabra 

27  virginiana 

28  clausa 

29  rigida 

30  serotina 

31  pungens 

32  banksiana 


Mexico 

Haploxylon  pines 

35  strobiformis 

36  cembroides 

Diploxylon  pines 
40  chihuahua 
53  engelmanni 

Northern  Eurasia 

Haploxylon  pines 

67  sibirica 

68  pumila 

Diploxylon  pines 

69  sylvestris 

Mediterranean  Region 

Haploxylon  pines 

70  cembra 

71  peuce 

Diploxylon  pines 

73  pinea 

74  wi^m 

75  heldreichii 

76  montana  i=mugo) 

77  halepensis 

78  brutia 
80  pinaster 

Eastern  Asia 

Haploxylon  pines 

82  koraiensis 

83  pentaphylla 

84  himekumatsu 
86  armandi 

91  wallichiana  (=griffithi) 

92  gerardiana 

Diploxylon  pines 
95  densiflora 
101  merkusii 
104  khasya 


178 


THE  REGENERATION  PROCESS 
OF  WHITEBARK  PINE 

Ward  W.  McCaughey 


Abstract — ^Whitebark  pine  (Pinus  albicaulis  Engelm.)  regener- 
ates similarly  to  European  and  Asian  stone  pines.  Our  knowl- 
edge of  this  process  is  essential  because  whitebark  pine  is  impor- 
tant to  the  North  American  grizzly  bear  {Ursus  arctos  horribilis), 
other  wildlife  species,  hydrology  of  high-elevation  ecosystems, 
and  esthetics.  This  paper  simimarizes  available  information  on 
the  whitebark  pine  regeneration  process,  beginning  from  bud  ini- 
tiation through  germination  and  seedling  survival.  Major  factors 
limiting  germination  and  regeneration  success  are  discussed.  We 
continue  to  gain  knowledge  about  the  regeneration  process  of 
whitebark  pine,  but  further  research  is  needed  to  fully  imder- 
stand  delayed  germination  mechanisms  and  habitat  require- 
ments for  optimimi  regeneration  success. 


Whitebark  pine  (Pinus  albicaulis  Engelm.)  is  an  impor- 
tant food  source  for  the  endangered  grizzly  bear  (Ursus 
arctos  horribilis),  red  squirrels  (Tamiasciurus  hudsonicus), 
the  Clark's  nutcracker  (Nucifraga  columbiana  Wilson), 
and  a  multitude  of  other  birds  and  mammals  (Hutchins 
and  Lanner  1982;  Kendall  1983;  Knight  and  others  1987). 
It  is  also  significant  as  a  hydrologic  stabilization  plant 
and  an  esthetic  feature  of  high-elevation  communities 
(Schmidt  and  McDonald  1990).  It  provides  important 
cover  for  wildlife  and  is  used  as  an  ornamental  for  land- 
scaping but  has  only  minor  significance  as  a  timber  pro- 
ducing species  (McCaughey  and  Schmidt  1990). 

General  information  is  available  and  is  summarized 
in  this  paper  on  the  natiu-al  regeneration  process  of  white- 
bark pine,  but  more  specific  facts  are  needed  for  efficient 
management  of  this  North  American  stone  pine  (Amo  and 
Hoff  1989;  McCaughey  and  Schmidt  1990).  This  paper 
describes  the  regeneration  process  beginning  firom  bud 
initiation  throiigh  germination  and  seedling  survival. 

CONE  AND  SEED  DEVELOPMENT 

Whitebark  pine  cone  and  seed  development  begins  vnth 
cone  initiation  and  ends  with  seed  matiiration.  Climatic 
conditions  influence  the  timing  of  development  and  matu- 
ration of  cone  initiation  and  seed  maturation. 

Cone  Initiation 

The  process  of  cone  initiation  and  development  of  stami- 
nate  and  ovulate  strobili  for  whitebark  pine  has  not  been 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Ward  W.  McCaughey  is  a  Research  Forester,  Intermountain  Research 
Station,  Forest  Service,  U.S.  Department  of  Agriculture,  Bozeman,  MT 
59717-0278. 


specifically  studied.  Mirov  (1967)  and  Krugman  and 
Jenkinson  (1974)  provided  generalized  descriptions  of  the 
initiation  process  for  the  genus  Pinus.  Female  and  male 
cone  initiation  of  whitebark  pine  probably  occiu-  during 
or  just  prior  to  winter  bud  formation  from  mid-August  to 
mid-September  (Schmidt  and  Lotan  1980). 

Cone  Development 

Female  and  male  buds  overwinter  and  begin  further 
development  and  growth  in  April  and  May  depending  on 
elevation.  In  1992,  staminate  cones  matured  and  shed 
pollen  during  Jxme  and  early  Jvdy  at  high  elevations 
(2,500-2,600  m)  in  southwestern  Montana  (McCaughey 
1992)  and  in  May  and  June  at  lower  elevations  (1,550- 
1,650  m)  in  northern  Idaho  (Hoff  1992).  Mature  male 
cones  are  about  1  cm  wide  by  1  cm  long  and  female  cones 
are  1  cm  wide  by  2  cm  long  during  early  development 
when  they  are  pollen  receptive.  After  pollination  female 
cones  grow  to  a  first-year  size  of  about  3  cm  long  by  2  cm 
wide.  Female  cones  remain  on  the  tree,  while  male  cones 
fall  off  after  pollen  dispersal.  Pollination  occurs  from  late 
Jime  to  late  Jvdy,  and  fertilization  occxirs  about  12  to  13 
months  later,  simultaneous  with  rapid  cone  enlargement 
during  June  and  July  (Krugman  and  Jenkinson  1974). 

Seed  Maturation 

Following  fertilization  the  embryo  develops  and  dif- 
ferentiates into  cotyledons,  plumule,  and  radicle  (Mirov 
1967).  Whitebark  seeds  reach  maturity  between  mid- 
August  and  mid-September  depending  on  elevation  and 
climatic  conditions  during  the  ripening  process  (Krugman 
and  Jenkinson  1974;  McCaughey,  in  press  a).  The  entire 
process  from  cone  initiation  to  cone  and  seed  matiuity 
takes  about  24  months. 

Cone  and  Seed  Insects  and  Diseases 

Whitebark  pine  cones  and  seeds  are  exposed  to  insects 
and  disease  that  reduce  cone  and  seed  siirvival  during  the 
2  years  of  development  (Edwards  1990).  Cone  and  seed 
insects  that  damage  whitebark  pine  are  cone  worms  (Dio- 
ryctria  spp.  and  Eucosma  spp.),  cone  beetles  (Conophtho- 
rus  spp.)  (Bartos  and  Gibson  1990),  and  midges  and  seed 
chalcids  (Megastigmus  spp.)  (Dewey  1989).  Siroccocus 
strobilinus  Preuss  is  a  seed-borne  disease  that  kills  white- 
bark pine  seedlings  in  niu-series  and  in  natxu-al  stands. 
Calocypha  fulgens  (Pers.)  Boud.  (anamorph  =  Geniculcden- 
dron  pyriforme  Salt),  referred  to  as  a  seed  or  cold  fungus, 
may  cause  preemergence  seed  loss  (Hoff  and  Hagle  1990). 
Pre-  and  postemergence  damping-off  diseases  (Fusarium 
spp.)  may  cause  extensive  mortality,  especially  in  slowly 
emerging  seedlings  (Landis  and  others  1990). 


179 


SEED  DISPERSAL 

Whitebark  pine  cones  are  well  suited  to  the  Clark's  nut- 
cracker, a  forest  bird  that  is  the  major  disperser  of  white- 
bark  pine  seeds  (Arno  and  Hoff  1989;  Hutchins  and 
Lanner  1982;  Lanner  1980;  Lanner  and  Vander  Wall  1980; 
McCaughey  and  Schmidt  1990;  Tomback  1982). 

Cone  Attributes 

Unlike  other  conifers  such  as  western  white  pine  (Pinus 
monticola  Dougl.)  and  limber  pine  {Pinus  flexilis  James), 
whose  cone  scales  fully  open  to  release  seed,  whitebark 
pine  cone  scales  only  open  partly.  The  scale  base  does  not 
break  away  from  the  cone  axis;  the  wingless  seeds  are 
held  firmly  in  place  yet  fully  exposed  and  easily  accessible 
to  nutcracker  extraction  (Eggers  1986;  Hutchins  and 
Lanner  1982).  Nutcrackers  have  strong,  pointed  bills  like 
other  species  of  the  family  Corvidae.  Upper  portions  of 
the  cone  scales  are  easily  broken  by  nutcrackers  along 
a  thin  fracture  zone  where  the  massive  apophysis  tapers 
to  a  thin  cross  section  beneath  the  seed-bearing  cavities 
(Lanner  1982).  A  nutcracker  can  store  over  100  white- 
bark pine  seeds  in  its  sublingual  pouch,  a  saclike  modifi- 
cation on  the  floor  of  the  mouth  (Bock  and  others  1973). 

Nutcracker  Caching 

Whitebark  pine  regeneration  depends  almost  entirely 
on  nutcracker  seed  selection  and  caching  habits.  A  nut- 
cracker appears  to  discriminate  between  good,  aborted, 
insect-infected,  or  diseased  seeds  by  rattling  each  seed 
in  its  bill  before  depositing  it  in  the  sublingual  pouch 
(Hutchins  and  Lanner  1982;  Tomback  1978;  Vander  Wall 
and  Balda  1977).  Whitebark  pine  seeds  are  dispersed  by 
nutcrackers  up  to  22  km  from  their  source.  Caches  are 
buried  2  to  4  cm  deep  with  1  to  25  or  more  seeds  per  cache 
(fig.  1)  (Tomback  1978;  Vander  Wall  and  Balda  1977). 
One  nutcracker  can  store  an  estimated  22,000  to  98,000 
whitebark  pine  seeds  each  year  when  seed  is  available 
(Hutchins  and  Lanner  1982;  Tomback  1978;  Vander  Wall 
and  Balda  1977).  Estimates  of  food  requirements  indicate 
that  nutcrackers  may  store  three  to  five  times  as  many 
seeds  as  needed  (Tomback  1983). 

Secondary  Dispersers 

Nutcrackers  compete  with  other  seed  consumers  for 
whitebark  pine  seeds.  Red  squirrels  harvest  cones  and 
seeds  from  mid-July  to  early  November,  greatly  reducing 
the  availability  of  seed  for  nutcrackers  (Eggers  1986; 
Hutchins  and  Lanner  1982;  Reinhart  and  Mattson  1990; 
Smith  1968).  Squirrels  store  cones  in  middens  and  seeds 
in  caches  on  the  forest  floor  (Hutchins  and  Lanner  1982; 
Reinhart  and  Mattson  1990). 

Grizzlies  obtain  whitebark  pine  seeds  primarily  from 
squirrel  middens  (Kendall  1983).  Germination  probabil- 
ity from  squirrel-cached  cones  and  seed  is  low  due  to  their 
deep  caching  habits  (>7  cm)  and  small  numbers  of  midden 
sites.  H5^ocotyl  growth  of  whitebark  pine  is  only  3  to  4  cm; 
emergence  of  cotyledons  above  the  soil  surface  in  a 


Figure  1 — Nutcracker  cache  site  with  five  white- 
bark pine  germinants  growing  on  a  litter  seedbed 
near  a  log. 


squirrel  midden  is  unlikely  (Lanner  1982).  Squirrels 
feed  on  the  cones  and  seed  during  the  winter  and  spring 
months  thus  constantly  disturbing  the  middens  (Lanner 
1982). 

Many  other  animals  feed  on  whitebark  pine  seed  and  are 
considered  secondary  dispersers  because  of  the  low  prob- 
ability of  germination  following  their  feeding  and  caching 
(Hutchins  and  Lanner  1982;  McCaughey  and  Schmidt 
1990).  Other  common  animals  that  harvest  whitebark 
pine  seed  either  from  the  cones  directly  or  indirectly  from 
the  ground  or  other  animal  caches  are:  birds — ^William's 
sapsucker  (Sphyrapicus  thyroideus),  hairy  woodpecker 
{Picoides  villosus),  white-headed  woodpecker  {P.  albolar- 
vatus),  mountain  chickadee  (Parus  gambeli),  white-breasted 
nuthatch  {Sitta  carolinensis),  Cassin's  finch  {Carpodacus 
cassinii),  red  crossbill  {Loxia  curvirostra),  pine  grosbeak 
(Pinicola  enucleator),  Steller's  jay  {Cyanocitta  stelleri), 
raven  (Corvus  corax),  and  red-breasted  nuthatch  {Sitta 
canadensis).  Mammals — chipmimks  {Eutamias  spp.), 
deer  mouse  {Peromyscus  maniculatus),  golden-mantled 
groimd  squirrel  {Spermophilus  lateralis),  southern  red- 
backed  vole  {Clethrionomys  gapperi),  chickaree  {Tamia- 
sciurus  douglasi)  (Hutchins  and  Lanner  1982;  McCaughey 
and  Schmidt  1990),  and  black  bears  {Ursus  americanus) 
(Craighead  and  others  1982;  Kendall  1983).  Secondary 
dispersers  not  only  reduce  the  seed  crop  but  limit  the 
availability  of  seed  to  nutcrackers  and  thus  the  probabil- 
ity of  regeneration. 

Usually  whitebark  pine  cone  crops  are  depleted  by 
cone  and  seed  consumers.  Hutchins  and  Lanner  (1982) 
described  a  3-year  period  when  no  whitebark  pine  cones 
fell  to  the  ground  other  than  by  animal  clipping.  In  con- 
trast, so  many  cones  were  produced  in  1989  that  many 
were  not  utilized  by  seed  consumers  and  fell  to  the  groimd 
(McCaughey  and  others  1990).  Grizzly  bears  were  ob- 
served foraging  on  these  cones  in  1990  (Reinhart  1990). 
The  1989  mast  year  may  have  reached  the  upper  limit  of 


180 


cone  production,  a  level  that  rarely  occiirs.  Wind  storms 
may  have  dislodged  the  whitebark  cones  making  them 
available  to  ground  foraging  animals. 

The  probability  of  germination  is  low  from  seeds  that 
have  fallen  to  the  ground  in  cones  or  been  dropped  by  ani- 
mals. All  whitebark  pine  seeds  sown  on  the  ground  are 
eaten  or  taken  by  seed  consxmiers  when  not  protected 
with  exclosures  (McCaughey  1990).  This  situation  oc- 
curred in  two  successive  seasons  when  seeds  were  sown 
during  field  germination  tests.  Eight  species  of  rodents 
were  trapped  on  site  dining  the  field  germination  test. 
The  two  most  abundant  species  were  the  deer  mouse  and 
the  southern  red-backed  vole  representing  54  and  23  per- 
cent of  the  total  popidation  trapped  (McCaughey  1990). 

GERMINATION 

Information  on  seed  storage  and  laboratory  germina- 
tion of  whitebark  pine  seeds,  under  a  variety  of  seed  treat- 
ment methods,  has  been  summarized  and  discussed  by 
several  authors  in  the  symposium  proceedings  "Whitebark 
Pine  Ecosystems:  Ecology  and  Management  of  a  High- 
Mountain  Resource"  (Schmidt  and  McDonald  1990).  The 
following  sections  extract  pertinent  information  from 
those  various  soiirces. 

Seed  Storage 

Whitebark  pine  seed  has  been  frozen  imder  environ- 
mentally controlled  conditions  for  up  to  20  years  (Schubert 
1954).  Seed  viability  decreased  over  that  time  period  from 
50  percent  to  3  percent.  Mirov  (1946)  found  that  viability 
of  whitebark  pine  seed  dropped  from  24  percent  at  time 
of  collection  to  17  percent  after  being  frozen  for  8  years, 
but  dropped  to  1  percent  after  11  years  of  storage.  Viabil- 
ity of  whitebark  pine  seed,  as  related  to  storage  time,  is 
dependent  on  seed  maturity  when  harvested,  seed  han- 
dling prior  to  storage,  and  methods  and  length  of  time 
of  stratification  for  germination  tests  (McCaughey  and 
Schmidt  1990). 

Controlled  Germination 

ViabiHty  of  whitebark  pine  seed,  even  iinder  controlled 
conditions,  is  highly  variable,  ranging  from  0  to  75  per- 
cent. One  procedure  for  germinating  whitebark  pine 
seeds  is  to  soak  seeds  in  running  tap  water  for  1  to  2  days 
and  stratify  moist  at  1  to  5  °C  for  90  to  120  days  in  plastic 
bags  (Krugman  and  Jenkinson  1974).  Stratification  time 
may  vary  by  seed  maturity.  Jacobs  and  Weaver  (1990a) 
found  that  1  month  of  stratification  was  sufficient  to  in- 
crease germination  from  5  percent  to  about  40  percent; 
longer  stratification  periods  (to  8  months)  did  not  improve 
germination.  The  Coevir  d'Alene  Nursery  in  Idaho, 
U.S.A.,  used  the  following  procedm-es  for  germination 
tests  for  western  white  pine  and  whitebark  pine: 

A.  Place  seed  in  nylon  mesh  bags. 

B.  Soak  seed  for  48  hours  in  ninning  tap  water.  Place 
nylon  mesh  bag  in  plastic  bag. 

C.  Stratify  for  100  days  at  1  to  2  °C.  Within  that  100- 
day  stratification  time  resoak  the  seed  for  1  hour  each 
week. 


D.  After  100  days,  remove  seed  from  stratification  and 
surface  dry. 

E.  Using  a  vacuum  seeder,  place  100  seeds  on  moist 
paper  towels  (Kimpak)  in  each  of  four  plastic  trays. 

F.  Place  trays  in  germinator  and  take  counts. 

Seedling  Morphology 

Whitebark  pine  germinants  have  thicker  stems  than 
those  of  associated  conifers.  Stem  diameters  range  from 
2  to  4  mm  for  whitebark  and  only  1  to  2  mm  for  lodgepole 
pine  {Pinus  contorta  var.  latifolia),  Engelmann  spruce 
{Picea  engelmannii  Parry  ex  Engelm.),  and  subalpine  fir 
iAhies  lasiocarpa  [Hook.]  Nutt.)  (McCaughey  1988).  Ger- 
minants produce  five  to  12  cotyledons  and  grow  to  heights 
of  3  to  5  cm  in  the  first  growing  season  (fig.  2)  (McCaughey 
and  Schmidt  1990).  Needles  vary  in  length  from  7  to  10 
cm  with  stomata  positioned  on  the  dorsal  and  ventral 
sides.  The  hypoderm  is  weak,  usually  one  cell  in  width; 
the  endodermis  has  an  outer  wall  that  is  strongly  thick- 
ened, and  there  is  only  one  fibrovascular  bundle  (Harlow 
1931). 

Whitebark  pine  trees  commonly  have  multiple  stems 
from  forking  of  a  single  stem  or  the  merging  of  several 
seedlings  from  a  nutcracker  cache.  Forking  occurs  on  a 
small  percentage  of  whitebark  pine  germinants.  Nearly 
7  percent  of  first-year  germinants  have  two  forks  and 
17  percent  have  three  or  more  forks  (McCaughey  1988; 
Weaver  and  Jacobs  1990).  Genotypic  analyses  of  multi- 
stemmed  clumps  show  that  58,  70,  and  83  percent,  of 
three  stands  sampled  in  Alberta,  Canada,  and  Wyoming, 
U.S.A.,  had  stems  of  mixed  genetic  origin  while  42,  30, 


Figure  2 — Single  whitebark  pine  seedling  germin- 
ating on  a  mineral  seedbed  and  showing  newly 
emerged  cotyledons. 


181 


and  17  had  arisen  by  branching  (Furnier  and  others 
1987;  Jacobs  and  Weaver  1990b;  Linhart  and  Tomback 
1985).  This  indicates  that  trees  of  mixed  genetic  origin 
had  merged  from  several  genetically  individual  seedlings 
from  the  same  cache. 

Whitebark  pine  germination  and  root  growth  are  affected 
by  early  season  soil  temperatures.  Seeds  germinate  when 
soil  temperatures  are  between  10  and  40  °C  with  an  opti- 
mum temperature  range  of  25  to  35  °C  (Jacobs  1989; 
Jacobs  and  Weaver  1990a).  Root  growth  occurs  between 
10  and  45  °C  with  the  optimum  temperature  range  being 
25  to  35  °C.  This  optimum  range  provides  conditions 
where  whitebark  roots  grow  5  to  15  mm  per  day  (Jacobs 
and  Weaver  1990a).  First-year  root  growth  ranges  be- 
tween 5  and  18  cm  for  nutcracker-cached  seedlings  grow- 
ing in  a  forest  environment  (McCaughey  1988). 

Period  of  Germination 

Germination  from  a  single  sowing  or  caching  of  white- 
bark pine  seeds  can  continue  throughout  the  growing  sea- 
son and  during  the  next  2  to  3  or  more  years  (McCaughey, 
in  press  b).  Because  nutcrackers  bury  seed  2  to  4  cm 
deep,  seeds  may  germinate  but  not  emerge  above  the  soil 
surface  for  several  days  after  germination  or  they  may  not 
emerge  at  all.  The  term  germinant  in  this  paper  refers 
to  an  emergent,  since  it  is  difficult  to  discern  if  a  seed  ger- 
minated yet  died  prior  to  emergence. 

Germination  Sites 

Whitebark  pine  germinates  on  a  variety  of  seedbeds, 
almost  entirely  dependent  on  the  choice  of  the  nutcracker. 
Whitebark  pine  grows  within  38  habitat-phase  combina- 
tions in  eastern  Idaho  and  western  Wyoming  (Steele  and 
others  1983)  and  44  in  Montana  (Pfister  and  others  1977). 
Typical  sites  for  Clark's  nutcracker  caches  are  well-drained 
and  moist  substrates,  bare  soil,  forest  litter,  gravel,  rubble, 
in  small  plants  and  logs,  in  cracks  and  fissures  on  exposed 
rock,  and  in  pumice  soils  (Hutchins  and  Lanner  1982; 
Tomback  1978,  1982;  Tomback  and  others  1990).  Regen- 
eration is  commonly  found  on  burned  litter  seedbeds  fol- 
lowing natural  and  prescribed  burns  (Morgan  and  Bimting 
1990;  Tomback  1986), 

Nutcrackers  cache  whitebark  seeds  on  all  aspects,  but 
the  majority  of  cache  sites  occur  on  southeast,  south, 
southwest,  and  west-southwest  aspects  (Tomback  and 
others  1990).  Seed  storage  in  south-aspect  windblown 
sites  probably  ensures  that  some  caches  are  snowfree  in 
winter  and  spring  so  the  nutcracker  can  retrieve  them. 
Even  though  fewer  seeds  are  cached  on  the  north  aspects, 
regeneration  densities  are  highest  on  north  aspects.  This 
may  be  due  to  less  successful  retrieval  by  the  nutcracker, 
or  it  may  be  a  result  of  favorable  environmental  condi- 
tions such  as  moisture  and  insolation  protection  during 
the  germination  period  (Tomback  and  others  1990).  The 
tallest  and  best  formed  whitebark  pine  trees  are  often 
found  in  high  basins  or  on  gentle  north  slopes  (Arno  and 


Hoff  1989).  Drought  and  insolation  mortality  of  seedlings 
on  south  slopes  may  contribute  to  the  disproportionate 
success  of  whitebark  regeneration  on  north  aspects. 

First-Year  Germination 

Whitebark  pine  seed  germinating  from  caches  by  the 
nutcracker  begins  about  mid-June  and  continues  through 
early  September.  The  number  of  whitebark  pine  seeds 
germinating  in  the  first  year  after  caching  varies  by  year, 
probably  in  response  to  timing  of  spring  snowmelt  and 
early  summer  rains  (fig.  3)  (McCaughey  1990).  For  ex- 
ample, in  1988  only  11.5  percent  of  buried  seeds  germi- 
nated, probably  due  to  below-average  precipitation  (82 
percent  of  normal).  In  1989  precipitation  was  near  normal 
and  germination  of  1988  sown  seed  was  nearly  34  percent 
(table  1)  (McCaughey  1990). 

First-year  germination  of  whitebark  pine  is  significant- 
ly affected  by  percent  shade  cover  (0,  25,  and  50  percent), 
seedbed  condition  (mineral,  litter,  and  burned),  and  sow- 
ing depth  (surface  sown  and  buried)  (table  1)  in  controlled 
field  studies  in  southwestern  Montana,  U.S.A.  (McCaughey 
1990).  Percent  germination  of  whitebark  pine  was  signifi- 
cantly ip  =  0.008)  higher  (20.4  percent)  under  a  50  percent 
shade  cover  than  with  no  shade  (16.7  percent)  (table  1) 
(McCaughey  1990).  Germination  of  whitebark  pine  seeds 
sown  in  1988  was  26  percent  on  mineral  seedbeds  and  sig- 
nificantly less  on  litter  and  burned  seedbeds,  14.5  and 
15.3  percent,  respectively  (McCaughey  1990).  Buried 
whitebark  pine  seeds  (2  to  4  cm  deep)  had  significantly 
higher  germination  in  1988  and  1989  (11.5  and  33.7  per- 
cent, respectively)  than  surface-sown  seeds  that  were  pro- 
tected from  seed  consumers  in  the  same  years  (1.8  and 
11.5  percent,  respectively)  (McCaughey  1990). 


50- 


?  40- 

a: 

UJ 
(D 

H  30" 
•z. 

UJ 

o 

cc 

lU 
Q. 

lU 
> 


20- 


^  10- 


3 
O 


Burled,  germinating  under  moist  conditions  - 1989 


JUNE 


n- 2.700  seeds 


Buried,  germinating  under  dry  conditions  - 1988 
n-  1,440  seeds 


JULY 


AUGUST    '  SEPTEMBER  '  OCTOBER 


Figure  3 — Cumulative  percent  germination  of  white- 
bark pine  germinants  from  buried  and  protected 
seeds,  sown  in  1987  and  1988  respectively,  germi- 
nating under  dry  (1988)  and  moist  (1989)  conditions. 
Data  collected  from  a  clearcut  at  2,530  m  elevation 
in  southwestern  Montana,  U.S.A. 


182 


Table  1 — Percent  first-year  germination  of  1987  and  1988  sown 

seeds  of  whitebark  pine  as  affected  by  shade  cover,  seed- 
bed condition,  and  sowing  depth.  All  treatments  excluded 
seed  consumers  allowing  for  surface  germination.  Data 
collected  from  a  clearcut  at  2,530  m  elevation  in  south- 
western Montana,  U.S.A. 


First-year  germination 

Factor 

1988 

1989 

Factor 

level 

mean 

mean 

-  Percent  

Shade  cover 

0 

^5.2  (a) 

16.7  (c) 

(percent) 

25 

7.9  (a) 

18.8  (cd) 

50 

6.8  (a) 

20.4  (d) 

Seedbed 

Mineral 

8.2  (a) 

26.0  (c) 

condition 

Litter 

5.1  (b) 

14.5  (d) 

Burned 

15.3  (d) 

Sowing  depth 

Surface 

1.8  (a) 

3.5  (c) 

Buried 

11.5  (b) 

33.7  (d) 

'Similar  and  dissimilar  letters  in  parentheses  within  a  column  for  a  factor 
represent  nonsignificant  and  significant  differences  respectively. 


Delayed  Germination 

Delayed  germinants  are  whitebark  pine  seeds  that  ger- 
minate two  or  more  seasons  after  being  sown  or  cached. 
The  European  stone  pine  (Pinus  cembra)  germinates  in 
the  first,  second,  and  third  year  after  caching  (Krugman 
and  Jenkinson  1974).  Whitebark  pine  seed  germinates 
at  least  3  years  following  sowing  (McCaughey  1990; 
McCaughey,  in  press  b;  McCaughey  and  Schmidt  1990). 
Although  there  were  absolute  value  differences,  the  effects 
of  shade  cover  and  seedbed  condition  became  nonsignifi- 
cant as  germination  occurred  over  a  3-year  period  in  a 
study  of  whitebark  germination  in  western  Montana, 
U.S.A.  (table  2)  (McCaughey,  in  press  b).  Germination 
from  buried  seed  (56  percent)  remained  significantly  higher 
than  for  surface-sown  seed  (7  percent)  3  years  after  sow- 
ing (McCavighey,  in  press  b). 

Dormancy  of  whitebark  pine  seed  is  caused  by  embryo 
underdevelopment,  physiological  embryo  dormancy,  im- 
perviousness  of  the  seed  coat  and  female  gametophyte 
tissue  to  oxygen  and  water  uptake,  and  possibly  by  depo- 
sition of  growth  inhibitors  to  the  embryo  by  female  game- 
tophyte tissue  (Pitel  1981;  Pitel  and  Wang  1980,  1990). 
Seeds  of  whitebark  pine  must  mature  in  a  short  growing 
season  in  high-elevation  forests,  and  if  climatic  conditions 
slow  the  maturation  process  nutcrackers  will  harvest  im- 
mature seed.  Nutcrackers  are  normally  not  concerned 
with  seed  maturity  and  will  also  harvest  seeds  early  in 
the  season  before  they  are  mature.  Embryo  vmderdevel- 
opment  can  be  overcome  by  exposing  imbibed  seeds  to 
20  °C  for  30  to  60  days  (Leadem  1985).  Clipping  the  seed- 
coat  is  a  method  to  overcome  physiological  barriers  to  ger- 
mination (Pitel  1981). 

Under  natural  conditions,  the  greatest  germination 
from  cached  whitebark  pine  seeds  occurs  in  the  second 


year  following  caching  due  to  delayed  dormancy  mecha- 
nisms (McCaughey  1990,  1992;  Tomback  1992).  For 
example,  germination  of  buried  whitebark  pine  seed  in- 
creased from  11  percent  in  the  first  year  following  sowing 
to  45  percent  in  the  second  year.  Germination  declined  to 
11  percent  in  the  third  year  foUovnng  sowing  (McCaughey, 
in  press  b). 

Limiting  Factors 

Germination  rates  for  whitebark  pine  are  highest  from 
mid-June  through  the  end  of  July  (McCaughey  1990). 
Mortality  of  first-year  seedlings  follows  the  same  pattern; 
the  highest  mortality  rates  occvir  when  germination  is 
highest  (fig.  4)  (McCaughey  1990).  Germination  exceeds 
mortality  up  to  the  first  of  August,  resulting  in  an  accu- 
mulation of  surviving  seedlings;  however,  germination 
and  mortality  rates  decrease  after  the  first  of  August. 
Germination  continues  from  early  August  until  the  first 
of  September  with  total  nimabers  of  survivors  remaining 
constant  because  mortality  equals  germination  (McCaughey 
1990).  Late  germinants  have  the  same  probability  of  svir- 
vival  as  early  germinants.  Factors  limiting  seedling  sur- 
vival include: 

Microsite  and  Biotic — Three  major  factors  are: 
(1)  insolation  (heat  scorching  of  seedling  stem  at  ground 
surface),  (2)  drought  (drying  out  of  seedling),  and  (3)  ani- 
mals (burial,  uprooting,  or  nipping),  specifically  pocket 
gophers  (Thomomys  talpoides)  (McCaughey  and  Schmidt 
1990).  Hutchins  and  Lanner  (1982)  observed  a  chipmunk 
uproot  and  consume  a  whitebark  pine  seedling.  Insola- 
tion mortality  of  whitebark  pine  was  higher  on  mineral, 
litter,  and  burned  seedbeds  on  nonshaded  plots  when 
compared  to  25  and  50  percent  shaded  plots  in  a  germina- 
tion test  in  southwestern  Montana,  U.S.A.  (McCaughey 
1990). 

Blister  Rust — The  introduced  disease,  blister  rust 
{Cronartium  ribicola),  poses  the  most  serious  threat  to 


Table  2 — Percent  germination'  of  1 987  sown  whitebark  pine  seed 
(buried  2  to  4  cm)  as  affected  by  shade  cover  and  seed- 
bed condition  for  each  of  the  first  3  years  following  sowing 
and  the  3-year  total.  Data  collected  from  a  clearcut  at 
2,530  m  elevation  in  southwestern  Montana,  U.S.A. 


Emergence 


Factor 


Factor 
level 


First 
year 


Second 
year 


Third 
year 


Three-year 
total 


Shade  cover 
(percent) 

Seedbed 
condition 


0 

25 
50 

Mineral 
Litter 


10 
13 
12 

13 
9 


Percent 

39 
51 
46 

40 
50 


5 
14 
15 

7 
15 


247  (a) 
61  (a) 
60  (a) 

51  (a) 
60  (a) 


'Percent  germination  for  each  year  is  based  on  the  number  of  seeds  that 
had  not  previously  emerged. 

^Similar  and  dissimilar  letters  in  parentheses  for  a  factor  represent  statisti- 
cally nonsignificant  and  significant  differences  respectively. 


183 


10 
8 

6  -I 


1988 


INSOLATION 
DROUGHT 


I 


JUNE 


JULY        AUGUST  SEPTEMBER 


1989 

■1  INSOLATION 

m  DROUGHT 

□  ANIMAL 

JUNE  1 

JULY     1  AUGUST  1  SEPTEMBER 

Figure  4 — Percent  first-year  mortality  of  whitebark 
pine  seedlings  by  causal  agent  over  time  for  1 987 
and  1 988  sown  seeds.  Data  collected  from  a  clear- 
cut  at  2,530  m  elevation  in  southwestern  Montana, 
U.S.A. 


The  Yellowstone  environment  is  poorly  suited  to  blister 
rust  infection  because  it  has  a  cool,  dry  climate  that  is 
only  marginally  suitable  for  blister  rust  teliospore  germi- 
nation (Krebill  1971). 

Mountain  Pine  Beetle — Mountain  pine  beetle  is  the 
most  damaging  insect  in  mature  stands  of  whitebark  pine 
(Arno  and  Hoff  1989;  McCaughey  and  Schmidt  1990). 
Large  numbers  of  whitebark  pine  are  killed  by  moimtain 
pine  beetle  when  epidemics  spread  into  whitebark  pine 
stands  from  the  lower  elevation  lodgepole  pine  zone.  These 
epidemics,  like  blister  rust,  create  conditions  where  suc- 
cession progresses  toward  dominance  by  shade-tolerant 
conifers  such  as  subalpine  fir  and  Engelmann  spruce  (Arno 
1986). 

Fire  Suppression — Whitebark  pine  is  serai  to  subal- 
pine fir  and  other  conifers,  and  periodic  fires  have  helped 
to  perpetuate  the  pine  (Arno  1986).  Past  fire  intervals 
in  whitebark  pine  stands  ranged  from  50  to  300  years  or 
more  depending  on  site  conditions  (Arno  1980).  Fires  oc- 
cur in  whitebark  pine  stands  only  under  severe  burning 
conditions  and  are  tj^ically  non-stand  replacing.  Fires 
sweep  across  large  areas  of  forest  reducing  competition 
from  shade-tolerant  conifers  (Arno  1976).  Nearly  90  years 
of  fire  suppression  have  reduced  the  annual  acreage  of 
whitebark  pine  forests  being  biu'ned,  creating  conditions 
where  shade-tolerant  species  have  taken  over  what  were 
once  serai  whitebark  pine  stands. 

Severe  fire  conditions  are  being  created  by  increased 
mortality  and  the  buildup  of  fuels  in  whitebark  pine  stands 
due  to  mountain  pine  beetle  and  blister  rust  epidemics. 
Stand  replacement  fires  will  be  the  result  of  this  unnatu- 
ral fuel  loading  and  will  eliminate  large  cone-producing 
stands  of  whitebark  pine. 


the  survival  of  whitebark  pine  in  parts  of  its  distribu- 
tional range.  Blister  rust  was  brought  to  western  North 
America  in  1910  on  a  boatload  of  blister  rust-infected 
eastern  white  pine  (Pinus  strobus  L.).  Whitebark  pine  is 
the  most  susceptible  to  blister  rust  infection  of  14  white 
pines  rated  by  Bingham  (1972).  Blister  rust  causes  exten- 
sive damage  and  mortality  in  whitebark  pine  stands  in 
moist  mountain  regions  in  northwestern  Montana,  north- 
ern Idaho,  and  the  Washington  Cascades  in  the  United 
States  (Arno  1986).  Damage  is  light  where  environmental 
conditions  are  somewhat  dry  and  cold,  and  there  is  a  de- 
crease in  low-elevation  sources  of  inoculimi  (Hoff  and 
Hagle  1990). 

Blister  rust-infected  trees  may  live  for  several  years 
before  mortality  occurs;  during  that  time  cone  production 
is  reduced  due  to  loss  of  upper  crowns.  Mortality  from 
rust  favors  succession  toward  dominance  by  whitebark 
pine's  shade-tolerant  associates  thus  reducing  site  poten- 
tial for  whitebark  pine  regeneration  and  survival  (Arno 
1986;  Arno  and  Hoff  1989;  McCaughey  and  Schmidt  1990). 

The  degree  of  blister  rust  infection  on  whitebark  pine 
decreases  southward  for  all  parts  of  its  range  (Hoff  and 
Hagle  1990;  Keane  and  others  1990).  A  major  population 
of  whitebark  pine  exists  in  the  Yellowstone  ecosystem  of 
southwestern  Montana  and  northwestern  Wyoming,  U.S.A. 


VEGETATIVE  REPRODUCTION 

Whitebark  pine  can  reproduce  vegetatively  through  lay- 
ering of  lower  branches  along  the  ground  surface.  In  the 
Mission  Range  of  western  Montana  vegetative  reproduc- 
tion was  observed  from  layering  of  shrublike  (krummholz) 
whitebark  pine  (Arno  1981;  Arno  and  Hoff  1989).  Although 
layering  is  possible,  the  vast  majority  of  reproduction  is 
from  seeds. 

SUMMARY 

Regeneration  processes  are  similar  for  whitebark  pine 
and  the  other  stone  pines  of  the  world.  The  initiation, 
development,  and  maturation  of  cones  and  seeds  for  white- 
bark pine  are  characteristic  of  five-needle  pines  in  general 
with  seasonal  timing  of  these  processes  varying  between 
species.  The  entire  process  from  cone  initiation  to  cone 
and  seed  maturity  takes  about  24  months.  Cone  and  seed 
development  are  most  affected  by  variations  in  climatic 
conditions  during  periods  of  reproductive  bud  formation, 
pollination,  and  growth  and  development.  Several  insects 
and  diseases  reduce  survival  of  cones  and  seeds  of  white- 
bark pine. 

Regeneration  of  whitebark  pine  depends  almost  exclu- 
sively on  the  seed  selection,  dispersal,  and  caching  habits 


184 


of  the  Clark's  nutcracker.  Other  stone  pines  of  the  world 
have  similar  mutualistic  relationships  with  bird  species 
for  seed  dispersal. 

Whitebark  pine  seeds  are  cached  by  Clark's  nutcrackers 
on  a  variety  of  sites  ranging  from  forest  litter  to  cracks 
and  fissures  in  rocks.  The  majority  of  regeneration  occurs 
on  south  or  west  aspects.  Seed  storage  in  south-aspect 
windblown  sites  ensures  that  some  caches  are  snow  free 
in  winter  and  spring  for  nutcracker  retrieval.  Regenera- 
tion densities  are  highest  on  north  aspects  even  though 
fewer  seeds  are  cached  there  because  of  favorable  micro- 
site  conditions  such  as  moistxu'e  and  insolation  protection. 

Germination  of  whitebark  pine  begins  about  mid-June 
and  continues  through  early  September  in  the  first  year 
following  caching.  Delayed  germination  occurs  for  up 
to  3  or  4  years  after  caching.  Total  germination  over  a 
3-year  period  is  highest  on  litter  seedbeds  that  have  50 
percent  shade  cover,  while  germination  is  lower  on  min- 
eral and  nonshaded  seedbeds  (McCaughey,  in  press  b). 
Dormancy  of  whitebark  pine  seed  is  caused  by  embryo 
underdevelopment,  physiological  embryo  dormancy,  and 
imperviousness  of  the  seed  coat  and  female  gametoph3^e 
tissue  to  oxygen  and  water  uptake. 

Factors  limiting  survival  of  whitebark  pine  seedlings 
are  microsite,  biotic,  disease,  insects,  and  fire.  Insolation 
and  drought  are  microsite-associated  factors  that  cause 
seedling  mortality.  Animals  such  as  pocket  gophers  bury, 
uproot,  or  nip  off  young  seedlings,  and  chipmunks  uproot 
and  consimie  whole  seedlings.  Deer  mice  and  southern 
red-backed  voles  and  other  small  mammals  consume  seed 
before  they  germinate.  Blister  rust  causes  mortality  or 
extensive  damage  of  the  cone-bearing  portions  of  the  tree. 
Whitebark  pine  is  killed  by  the  mountain  pine  beetle 
when  epidemics  spread  into  whitebark  pine  stands  from 
the  lower  elevation  lodgepole  pine  zone.  Fire  suppression 
creates  conditions  where  shade-tolerant  species  such  as 
Engelmann  spruce  and  subalpine  fir,  normally  killed  by 
light  imderburns,  take  over  through  succession  what  were 
once  serai  whitebark  pine  stands.  Severe  fire  conditions 
are  being  created  by  the  buildup  of  fuels  in  whitebark 
pine  stands  due  to  increased  mortality. 

We  are  beginning  to  piece  together  the  puzzle  of  the  re- 
generation process  of  whitebark  pine.  Survival  of  wildlife 
species  such  as  the  grizzly  is  threatened  due  to  a  reduc- 
tion or  loss  of  localized  populations  of  whitebark  pine 
caused  by  global  climate  change,  introduced  disease,  in- 
sects, and  succession  fi-om  fire  exclusion.  Future  research 
should  define  microsite  and  biotic  factors  influencing  ger- 
mination and  long-term  survival  of  seedlings.  Informa- 
tion on  regeneration  processes  from  other  stone  pine  spe- 
cies from  around  the  world  is  helping  identify  key  areas 
for  research  on  whitebark  pine.  For  example,  little  infor- 
mation is  available  on  the  elevational  distribution  of 
whitebark  and  which  habitats  are  best  suited  for  regen- 
eration. Cembra  pine  {Pinus  cembra)  grows  in  the  Swiss 
Alps  fi-om  valley  bottoms  to  timberline  when  management 
actions  reduce  intertree  competition.  Whitebark  pine 
management  wiU  improve  from  increased  knowledge  of 
regeneration  processes  of  all  the  stone  pines. 


REFERENCES 

Amo,  S.  F.  1976.  The  historical  role  of  fire  on  the  Bitter- 
root  National  Forest.  Res.  Pap.  INT-187.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Forest  and  Range  Experiment  Station.  30  p. 

Arno,  S.  F.  1980.  Forest  fire  history  in  the  Northern 
Rockies.  Journal  of  Forestry.  78(8):  460-465. 

Arno,  S.  F.  1981.  [Unpublished  observations].  Missoula, 
MT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermoimtain  Fire  Sciences  Laboratory. 

Amo,  S.  F.  1986.  Whitebark  pine  cone  crops  -  a  diminish- 
ing source  of  wildlife  food?  Western  Journal  of  Applied 
Forestry.  1(3):  92-94. 

Arno,  S.  F.;  Hoff,  R.  J.  1989.  Silvics  of  whitebark  pine 
(Pinus  albicaulis).  Gen.  Tech.  Rep.  INT-253.  Ogden, 
UT:  U.S.  Department  of  Agriculture,  Forest  Service, 
Intermoimtain  Research  Station,  lip. 

Bartos,  D.  L.;  Gibson,  K.  E.  1990.  Insects  of  whitebark 
pine  with  emphasis  on  mountain  pine  beetle.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain  re- 
source; 1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
171-178. 

Bingham,  R.  T.  1972.  Taxonomy,  crossability,  and  relative 
blister  rust  resistance  of  5-needled  white  pines.  In:  Biol- 
ogy of  rust  resistance  in  forest  trees:  proceedings  of  a 
NATO-IUFRO  advance  study  institute;  1969  August 
17-24;  Moscow,  ID.  Misc.  Publ.  1221.  Washington,  DC: 
U.S.  Department  of  Agriculture:  271-280. 

Bock,  W.  J.;  Balda,  R.  P.;  Vander  Wall,  S.  B.  1973.  Mor- 
phology of  the  sublingual  pouch  and  tongue  muscula- 
ture in  Clark's  nutcracker.  Auk.  90:  491-519. 

Craighead,  J.  J.;  Sumner,  J.  S.;  Scaggs,  G.  B.  1982.  A  de- 
finitive system  for  analysis  of  grizzly  bear  habitat  and 
other  wilderness  resources.  Wildlands  Inst.  Monogr.  1. 
Missoula,  MT:  University  of  Montana.  279  p. 

Dewey,  Jerald  E.  1989.  [Personal  communication  to  Dale 
Bartos  of  Intermountain  Research  Station,  Logan,  UT]. 
March  27.  Missoula,  MT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Northern  Region. 

Edwards,  D.  G.  W.  1990.  Cone  prediction,  collection  and 
processing.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
78-102. 

Eggers,  D.  E.  1986.  Management  of  whitebark  pine  as  po- 
tential grizzly  bear  habitat.  In:  Contreras,  G.  P.;  Evans, 
K.  E.,  comps.  Proceedings — grizzly  bear  habitat  sympo- 
sium; 1985  April  30-May  2;  Missoula,  MT.  Gen.  Tech. 
Rep.  INT-207.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
170-175. 


185 


Furnier,  G.;  Knowles,  P.;  Clyde,  M.;  Dancik,  B.  1987.  Ef- 
fects of  avian  seed  dispersal  on  the  genetic  structure 
of  whitebark  pine  populations.  Evolution.  41:  607-612. 

Harlow,  W.  M.  1931.  The  identification  of  the  pines  of  the 
United  States,  native  and  introduced,  by  needle  struc- 
ture. Tech.  Publ.  32.  Syracuse,  NY:  New  York  State 
College  of  Forestry.  21  p. 

Hoff,  R.  J.  1992.  Unpublished  data  on  file  at:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station,  Forestry  Sciences  Laboratory,  Moscow, 
ID. 

Hoff,  R.  J.;  Hagle,  S.  1990.  Diseases  of  whitebark  pine 
with  special  emphasis  on  white  pine  blister  rust.  In: 
Schmidt,  Wyman  C;  McDonald,  Kathy  J.,  comps. 
Proceedings — symposium  on  whitebark  pine  ecosys- 
tems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station: 
179-190. 

Hutchins,  H.  E.;  Lanner,  R.  M.  1982.  The  central  role  of 
Clark's  nutcracker  in  the  dispersal  and  establishment 
of  whitebark  pine.  Oecologia.  55:  192-201. 

Jacobs,  J.  S.  1989.  Temperature  and  light  effects  on  seed- 
ling performance  of  Pinus  albicaulis.  Bozeman,  MT: 
Montana  State  University.  39  p.  Thesis. 

Jacobs,  J.;  Weaver,  T.  1990a.  Effects  of  temperature  and 
temperature  preconditioning  on  seedling  performance 
of  whitebark  pine.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  white- 
bark pine  ecosystems:  ecology  and  management  of  a 
high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain 
Research  Station:  134-139. 

Jacobs,  J.;  Weaver,  T.  1990b.  Occurrence  of  multiple 
stems  in  whitebark  pine.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-moimtain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  156-159. 

Keane,  R.  E.;  Arno,  S.  F.;  Brown,  J.  K.;  Tomback,  D.  F. 
1990.  Simulating  disturbances  and  conifer  succession 
in  whitebark  pine  forests.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposium 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  274-288. 

Kendall,  K.  C.  1983.  Use  of  pine  nuts  by  grizzly  and  black 
bears  in  the  Yellowstone  area.  In:  International  Confer- 
ence on  Bear  Research  and  Management.  Madison,  WI: 
International  Bears  Association.  5:  166-173. 

Knight,  R.  R.;  Blanchard,  B.  M.;  Mattson,  D.  J.  1987. 
Yellowstone  grizzly  bear  investigation:  report  of  the 
interagency  study  team,  1987.  Bozeman,  MT:  U.S.  De- 
partment of  the  Interior,  National  Park  Service.  80  p. 


Krebill,  R.  G.  1971.  Conditions  for  teliospore  germination 
in  Yellowstone  National  Park  environment.  Unpub- 
lished report  on  file  at:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Research  Station, 
Ogden,  UT. 

Krugman,  S.  L.;  Jenkinson,  J.  L.  1974.  Pinus  L.  pine. 
In:  Schopmeyer,  C.  S.,  ed.  Seeds  of  woody  plants  in  the 
United  States.  Agric.  Handb.  450.  Washington,  DC: 
U.S.  Department  of  Agriculture,  Forest  Service:  598-638. 

Landis,  T.  D.;  Tinus,  R.  W.;  McDonald,  S.  E.;  Barnett, 
J.  P.  1990.  The  biological  component:  nursery  pests  and 
mycorrhizae.  Vol.  5.  The  container  tree  nursery  manual. 
Agric.  Handb.  674.  Washington,  DC:  U.S.  Department 
of  Agriculture,  Forest  Service.  171  p. 

Lanner,  R.  M.  1980.  Avian  seed  dispersal  as  a  factor  in 
the  ecology  and  evolution  of  limber  and  whitebark 
pines.  In:  Sixth  North  American  forest  biology  work- 
shop proceedings;  1980  August;  Edmonton,  AB. 
Edmonton,  AB:  University  of  Alberta:  14-48. 

Lanner,  R.  M.  1982.  Adaptations  of  whitebark  pine  for 
seed  dispersal  by  Clark's  nutcracker.  Canadian  Journal 
of  Forest  Research.  12(2):  391-402. 

Lanner,  R.  M.;  Vander  Wall,  S.  B.  1980.  Dispersal  of  lim- 
ber pine  seed  by  Clark's  nutcracker.  Journal  of  For- 
estry. 78(10):  637-639. 

Leadem,  C.  L.  1985.  Seed  dormancy  in  three  Pinus  species 
of  the  Inland  Mountain  West.  In:  Schmidt,  Wyman  C, 
comp.  Proceedings — future  forests  of  the  Moimtain 
West:  a  stand  culture  symposium;  1986  September  29- 
October  3;  Missoula,  MT.  Gen.  Tech.  Rep.  INT-243. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Research  Station:  117-124. 

Linhart,  Y.;  Tomback,  D.  F.  1985.  Seed  dispersal  by  nut- 
crackers causes  multi-trunk  growth  form  in  pines. 
Oecologia.  67:  107-110. 

McCaughey,  W.  W.  1990.  Biotic  and  microsite  factors 
affecting  Pinus  albicaulis  establishment  and  survival. 
Bozeman,  MT:  Montana  State  University.  78  p. 
Dissertation. 

McCaughey,  W.  W.  1988.  Unpublished  data  on  file  at: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  Forestry  Sciences  Labora- 
tory, Bozeman,  MT. 

McCaughey,  W.  W.  1992.  Unpublished  data  on  file  at: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station,  Forestry  Sciences  Labora- 
tory, Bozeman,  MT. 

McCaughey,  W.  W.  [In  press  a].  Seasonal  maturation  of 
whitebark  pine  seed  in  the  Greater  Yellowstone  ecosys- 
tem. In:  Schullery,  Paul;  Despain,  Don,  comps.  Proceed- 
ings of  the  1st  biennial  scientific  conference  on  the 
Greater  Yellowstone  Ecosystem — plants  and  their  envi- 
ronments; 1991  September  16-17;  Mammoth  Hot 
Springs,  Yellowstone  National  Park,  WY.  Tech.  Rep. 
USDI  National  Park  Service. 

McCaughey,  W.  W.  [In  press  b].  Delayed  germination  and 
seedling  emergence  of  Pinus  albicaulis  in  a  high  eleva- 
tion clearcut  in  Montana,  U.S.A.  In:  Edwards,  D.  G., 
comp.  Proceedings — symposium:  seed  dormancy  and 
barriers  to  germination;  1991  April  23-26;  lUFRO 


186 


project  group  P2. 04-00  meeting  on  seed  problems; 
Victoria,  BC.  Victoria,  BC:  Forestry  Canada. 

McCaughey,  W.  W.;  Amo,  S.  F.;  Tomback,  D.  F.  1990. 
Unpublished  personal  observations  from  Henderson 
Mountain  near  Cooke  City,  MT,  U.S.A..  On  file  at:  U.S. 
Department  of  Agriciilture,  Forest  Service,  Intermoun- 
tain  Research  Station,  Forestry  Sciences  Laboratory, 
Bozeman,  MT. 

McCaughey,  W.  W.;  Schmidt,  W.  C.  1990.  Autecology  of 
whitebark  pine.  In:  Schmidt,  Wjonan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
mountain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermoimtain  Research 
Station:  85-96. 

Mirov,  N.  T.  1946.  Viability  of  pine  seed  after  prolonged 
cold  storage.  Journal  of  Forestry.  44(3):  193-195. 

Mirov,  N.  T.  1967.  The  genus  Pinus.  New  York:  The  Ronald 
Press.  602  p. 

Morgan,  Penny;  Bunting,  Stephen  C.  1990.  Fire  effects 
in  whitebark  pine  forests.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposiimi 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain  Research  Station:  166-170. 

Pfister,  R.  D.;  Kovalchik,  B.  L.;  Arno,  S.  F.;  Presby,  R.  C. 
1977.  Forest  habitat  types  of  Montana.  Gen.  Tech.  Rep. 
INT-34.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Forest  and  Range  Ex- 
periment Station.  174  p. 

Pitel,  J.  A.  1981.  Improved  germination  of  whitebark 
pine  {Pinus  alhicaulis  Engelm.)  seeds.  Chalk  River,  ON: 
Canadian  Forestry  Service,  Petawawa  National  For- 
estry Institute;  unpublished  report.  5  p. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1980.  A  prehminary  study  of 
dormancy  in  Pinus  albicaulis  seeds.  Bi-Monthly  Re- 
search Notes.  36(1):  4-5. 

Pitel,  J.  A.;  Wang,  B.  S.  P.  1990.  Physical  and  chemical 
treatments  to  improve  germination  of  whitebark  pine 
seeds.  In:  Schmidt,  Wyman  C;  McDonald,  Kathy  J., 
comps.  Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-movmtain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agriciil- 
ture,  Forest  Service,  Intermoimtain  Research  Station: 
130-133. 

Reinhart,  D.  P.  1990.  [Personal  commmiication].  August. 
Bozeman,  MT:  U.S.  Department  of  the  Interior; 
Montana  State  University,  Interagency  Grizzly  Bear 
Project. 

Reinhart,  D.  P.;  Mattson,  D.  J.  1990.  Red  sqiiirrels  in  the 
whitebark  zone.  In:  Schmidt,  Wyman  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  whitebark 
pine  ecosystems:  ecology  and  management  of  a  high- 
moxmtain  resource;  1989  March  29-31;  Bozeman,  MT. 
Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Department 
of  Agriculture,  Forest  Service,  Intermoimtain  Research 
Station:  256-263. 


Schmidt,  W.  C;  Lotan,  J.  E.  1980.  Phenology  of  common 
forest  flora  of  the  Northern  Rockies— 1928  to  1937.  Res. 
Pap.  INT-259.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Forest  and  Range 
Experiment  Station.  20  p. 

Schmidt,  W.  C;  McDonald,  K.  J.,  comps.  1990. 
Proceedings — symposium  on  whitebark  pine  eco- 
systems: ecology  and  management  of  a  high-mountain 
resource;  1989  March  29-31;  Bozeman,  MT.  Gen.  Tech. 
Rep.  INT-270.  Ogden,  UT:  U.S.  Department  of  Agricul- 
ture, Forest  Service,  Intermountain  Research  Station. 
386  p. 

Schubert,  G.  H.  1954.  Viability  of  various  coniferous  seeds 
after  cold  storage.  Journal  of  Forestry.  52:  446-447. 

Smith,  C.  C.  1968.  The  adaptive  nature  of  social  organiza- 
tion in  the  genus  of  tree  squirrels  Tamiasciurus .  Eco- 
logical Monographs.  38:  31-63. 

Steele,  R.;  Cooper,  S.  V.;  Ondov,  D.  M.;  Roberts,  D.  W.; 
Pfister,  R.  D.  1983.  Forest  habitat  types  of  eastern 
Idaho-western  Wyoming.  Gen.  Tech.  Rep.  INT-144. 
Ogden,  UT:  U.S.  Department  of  Agriculture,  Forest 
Service,  Intermountain  Forest  and  Range  Experiment 
Station.  122  p. 

Tomback,  D.  F.  1978.  Foraging  strategies  of  Clark's  nut- 
cracker. Living  Bird.  16(1977):  123-160. 

Tomback,  D.  F.  1982.  Dispersal  of  whitebark  pine  seeds 
by  Clark's  nutcracker:  a  mutualism  hypothesis.  Journal 
of  Animal  Ecology.  51(2):  451-467. 

Tomback,  D.  F.  1983.  Nutcrackers  and  pines:  coevolution 
or  coadaptation?  In:  Nitecki,  M.  H.,  ed.  Coevolution. 
Chicago  and  London:  University  of  Chicago  Press: 
179-223. 

Tomback,  D.  F.  1986.  Post-fire  regeneration  of  krummholz 
whitebark  pine:  a  consequence  of  nutcracker  seed  cach- 
ing. Madrono.  33(2):  100-110. 

Tomback,  D.  F.  1992.  Unpublished  data  on  file  at:  Depart- 
ment of  Biology,  University  of  Colorado  at  Denver, 
Denver,  CO. 

Tomback,  D.  F.;  Hoffmann,  L.  A.;  Sund,  S.  K.  1990.  Co- 
evolution  of  whitebark  pine  and  nutcrackers:  implica- 
tions for  forest  regeneration.  In:  Schmidt,  Wyman  C; 
McDonald,  Kathy  J.,  comps.  Proceedings — symposiima 
on  whitebark  pine  ecosystems:  ecology  and  manage- 
ment of  a  high-mountain  resource;  1989  March  29-31; 
Bozeman,  MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT: 
U.S.  Department  of  Agriculture,  Forest  Service,  Inter- 
mountain Research  Station:  118-129. 

Vander  Wall,  S.  B.;  Balda,  R.  P.  1977.  Coadaptations  of 
the  Clark's  nutcracker  and  the  pinon  pine  for  efficient 
seed  harvest  and  dispersal.  Ecological  Monographs. 
47:  89-111. 

Weaver,  T.;  Jacobs,  J.  1990.  Occurrence  of  multiple  stems 
in  whitebark  pine.  In:  Schmidt,  Wjonan  C;  McDonald, 
Kathy  J.,  comps.  Proceedings — symposium  on  white- 
bark pine  ecosystems:  ecology  and  management  of  a 
high-mountain  resource;  1989  March  29-31;  Bozeman, 
MT.  Gen.  Tech.  Rep.  INT-270.  Ogden,  UT:  U.S.  Depart- 
ment of  Agriculture,  Forest  Service,  Intermountain  Re- 
search Station:  156-159. 


187 


JAPANESE  STONE  PINE  CONE 
PRODUCTION  ESTIMATED  FROM  CONE 
SCARS,  MOUNT  KISOKOMAGATAKE, 
CENTRAL  JAPANESE  ALPS 

Ikuko  Nakashinden 


Abstract — The  past  cone  production  of  the  Japanese  stone  pine 
(Pinus  pumila)  on  Mount  Kisokomagatake  was  reconstructed 
from  cone  scars.  The  main  cone  production  was  by  larger  indi- 
viduals in  the  lower  part  of  the  scrub  zone;  it  was  greater  than 
that  in  the  upper  part  where  growth  and  crop  are  controlled 
mainly  by  strong  winter  winds  and  snow. 


The  Japanese  stone  pine  {Pinus  pumila  Regal)  is  a 
genetic  dwarf  pine  distributed  over  the  eastern  part  of 
Siberia  and  China,  the  northern  part  of  Korea,  and  Japan. 
In  Japan,  most  of  its  habitats  are  located  on  high  moun- 
tains. The  climax  scrub  zone  is  formed  above  the  timber- 
line  of  subalpine  coniferous  forests.  At  the  southern  mar- 
gin of  its  range  (Chubu  District)  the  scrub  zone  occupies 
the  altitudinal  zone  from  1,961  to  3,192  m  (Yanagimachi 
and  Ohmori  1991). 

The  Japanese  stone  pine  propagates  in  two  different 
ways:  one  by  the  expansion  of  shoots  and  the  other  by 
seeds.  The  latter  is  the  most  important  for  spatial  ex- 
pansion (Okitsu  1990).  Seeds  are  dispersed  by  seed- 
storing  corvids  {Nucifraga  caryocatactes)  (Hayashida 
1989;  Tomback  1990).  Studies  concerning  pine  cone 
production  have  begun  recently.  Okitsu  and  Mizoguchi 
(1990,  1991)  evaluated  pine  cone  productivity  based  on 
the  individual  growth  sizes  and  scrub  sizes.  In  addition, 
studies  of  the  relationship  between  environment  of 
the  habitats  and  cone  production  have  commenced 
(Nakashinden  1990).  These  studies  are  limited  to  1  year's 
cone  production  data.  It  is  notable  that  past  cone  produc- 
tion can  be  estimated  from  cone  scars.  This  was  done  on 
whitebark  pine  {Pinus  albicaulis)  in  the  United  States 
(Morgan  and  Bimting  1992;  Weaver  and  Forcella  1986). 
Saito  and  others  (1989)  reported  that  cone  scars  are  also 
recognizable  in  the  Japanese  stone  pine  in  Hokkaido,  but 
little  information  about  it  was  given  in  their  report. 

This  author  developed  the  method  to  reveal  the  yearly 
change  in  cone  production  for  the  Japanese  stone  pine  and 
succeeded  in  reconstructing  pine  cone  production  over  the 
past  15  years  (Nakashinden  1991).  Using  this  cone  scar 
method,  chronological  pine  cone  production  of  the  Japa- 
nese stone  pine  is  described  and  the  relationship  to  the 
environment  of  its  habitats  is  examined. 


STUDY  AREA  AND  METHODS 

The  study  area  is  located  on  Moimt  Kisokomagatake, 
whose  simimit  is  2,956  m,  in  the  Central  Japanese  Alps, 
in  Chubu  District  (fig.  1).  This  location  is  one  of  the 
southernmost  parts  of  the  Japeinese  stone  pine 
distribution. 

The  habitats  examined  are  located  on  the  mountain 
slopes  between  2,600  and  2,950  m  elevation,  including 
the  summits  of  Mount  Kisokomagatake  and  Mount 
Chausudake  (2,652  m).  In  this  altitudinal  zone,  the 
Japanese  stone  pine  is  dominant,  forming  its  own  vege- 
tation zone  above  the  subalpine  forest  of  Abies  veitchii, 
Tsuga  diversifolia,  Betula  ermanii,  and  Alnus 
maximowiczii. 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Ikuko  Nakashinden  is  Graduate  Student,  Department  of  Geography, 
University  of  Tokyo,  Bunkyo-ku,  Tokyo  113,  Japan. 


Figure  1— Location  of  Mount  Kisokomagatake. 


188 


Pine  Cone  Scars 

The  positions  of  pine  cones  of  the  Japanese  stone  pine 
are  restricted  to  the  first  nodes  of  the  annual  shoots. 
They  matiire  in  the  second  year  after  fruiting.  In 
September,  nutcrackers  harvest  cones  of  the  Japanese 
stone  pine  as  a  food  source.  When  the  pine  cone  is  picked, 
a  scar  of  the  cone  remains  on  the  stem  (Nakashinden 
1991;  Saito  and  others  1989).  Each  scar  represents  one 
cone.  From  these  scars,  the  number  of  pine  cones  can  be 
calciilated.  The  production  can  be  determined  from  the 
location  of  the  nodes  on  the  stems.  This  cone  scar  method 
was  used  on  whitebark  pine  in  the  United  States,  and  the 
scars  could  be  measured  for  6  to  8  years  (Weaver  and 
Forcella  1986),  or  6  to  12  years  (Morgan  and  Bunting 
1992).  For  the  Japanese  stone  pine,  the  scars  could  be 
counted  for  15  to  20  years  (Nakashinden  1991). 

Using  this  method,  32  plots  (1  m^)  at  different  altitudes 
from  2,600  m  near  the  forest  hne  to  2,950  m  near  the 
summit  were  investigated.  Ten  stems  in  each  plot  were 
selected  for  counting  pine  cone  scars  and  for  the  measure- 
ment of  individual  stem  growth  for  the  15  years  preceding 
1990. 

Distribution,  Geomorphology,  and 
Snow 

Using  color  aerial  photographs  and  field  observation, 
the  Japanese  stone  pine  scrubs  in  the  study  area  were 
delineated  and  mapped.  Some  topographical  maps  were 
made  during  the  field  survey  in  order  to  examine  the  mi- 
croscale  landform  beneath  the  scrubs.  The  snow  melting 
status  and  snow  melting  period  near  the  simimit  of 
Moxmt  Kisokomagatake  was  observed  and  recorded  on 
photographs  in  each  spring  season  from  1989  to  1991. 


250  - 

200  - 


0) 

o 

0 


150  - 


100  - 


50- 


r— T 

76  77  78  79  80  81  82 

Year 


II   I  '  I   I  I  I 

83  84  85  86  87  88  89  90 


Figure  2 — Annual  cone  crops  of  Japanese  stone 
pine  (1976-90).  Cone  crops  were  reconstructed 
from  cone  scars  on  the  stems.  Total  number  from 
32  plots  is  shown  for  each  year. 


V) 

o 
o 
E 

C 

o 
O 


4.0 


3.0 


2.0 


1.0 


1983 


J  L 


J  L 


76 


78 


80 


82  84 

Year 


86  88 


90 


Figure  3 — Mean  annual  stem  elongation  of 
Japanese  stone  pine  (1976-90).  Ten  stems 
each  in  32  plots  were  measured. 


CONE  PRODUCTION  OVER  15  YEARS 

The  past  15  years  of  pine  cone  production  was  recon- 
structed on  the  study  area  (fig.  2).  Mast  years  were 
recognized  in  1981,  1984,  1988,  and  1989,  showing  a 
3-  or  4-year  periodicity  of  cone  production.  The  lowest 
production  during  the  past  15  years  occurred  in  1982  and 
1983.  In  1983,  the  annual  shoot  length  was  also  the  low- 
est during  the  15-year  period  (fig.  3).  The  year  1983  is 
judged  to  be  one  of  the  worst  years  for  both  cone  crop  and 
growth. 

On  the  individual  plots,  both  mast  and  fail  years  oc- 
curred regardless  of  their  environmental  differences. 
For  the  mast  years,  68  percent  of  the  32  plots  showed 
high  production  in  1981,  and  in  1984  it  was  65  percent. 
For  the  fail  years,  59  percent  of  the  plots  failed  in  cone 
production  in  1982,  and  in  1983  it  was  63  percent.  These 
occurrences  were  dependent  on  the  climate  and  not  on 
the  growth  forms  or  the  location  of  scrubs. 

DISTRffiUTION  PATTERNS  BY 
ALTITUDE 

On  Mount  Kisokomagatake,  the  Japanese  stone  pine 
scrubs  dominate  and  form  a  vertical  zone  above  the  subal- 
pine  forest  at  an  elevation  between  2,600  and  2,950  m. 
However,  the  distribution  patterns  were  characterized  by 
the  altitude.  On  the  upper  part  from  2,850  up  to  2,950  m, 
near  the  summit  of  Moimt  Kisokomagatake,  the  distribu- 
tion pattern  of  the  scrubs  shows  a  stripe  pattern  (fig.  4). 

The  scrubs  are  fragmented  by  huge  granite  rocks,  tors, 
stone  bank  terraces,  trails,  and  alpine  plant  communities 
such  as  Arctous  alpinus  vai.Japonicus,  Diapensia  lap- 
ponia  var.  obovata,  Oxytropis  japonica  Maxim,  Leonto- 
podium  shinanense  Kitam,  Gentiana  algida  Pall.,  and 
Loiseleuria  procumhens  Desv.  The  lower  part  between 
2,850  and  2,600  m,  near  the  boimdary  to  the  subalpine 
forest  shows  an  extensively  spreading  scrub,  covering  the 
slopes  entirely  from  the  ridges  down  to  the  forest  limit 
(fig.  5). 


189 


Figure  4 — Distribution  pattern  of  Japanese  stone 
pine  scrubs  in  tine  upper  part  of  the  Mount 
Kisokomagatake  study  area.  Distribution  is  frag- 
mented by  huge  rocks,  tors,  stone  bank  terraces, 
trails,  and  alpine  plant  communities. 


GROWTH  FORM  AND  CONE 
PRODUCTION  RELATED  TO 
ALTITUDE 

Growth  form  and  pine  cone  production  are  ditTerent  be- 
tween the  upper  and  the  lower  part  of  the  Japanese  stone 
pine  scrub  zone  (table  1).  The  mean  scrub  height,  stem 
length,  and  diameter  at  the  stem  base  in  the  lower  part 
are  almost  twice  as  large  as  those  in  the  upper  part.  The 
mean  pine  cone  production  per  stem  is  1.5  times  more 
than  in  the  upper  part.  Although  the  niunber  of  stems 
per  square  meter  in  the  upper  part  was  twice  that  of  the 
lower  part,  the  Japanese  stone  pine  scrubs  in  the  lower 
part  were  larger  in  size  and  produced  more  cones  than  the 
scrubs  in  the  upper  part. 


DISCUSSION  AND  CONCLUSIONS 

On  Moimt  Kisokomagatake,  the  mean  summer  tem- 
perature (July  and  August)  was  10  to  12  °C  at  2,600  m 
(Koizumi  1974)  and  10  °C  (Nakashinden  1990,  unpub- 
lished) at  2,850  m.  Concerning  the  lower  parts  of  the 
Japanese  stone  pine  scrub  zone,  the  thermal  conditions 
indicate  that  the  Japanese  stone  pine  scrub  zone  is  in- 
cluded in  the  thermal  subalpine  zone  (Yanagimachi  and 
Ohomori  1991).  There,  the  Japanese  stone  pine  scrubs 
are  considered  to  grow  and  produce  cones  at  the  same 
rates  as  in  the  subalpine  zone. 

The  scrubs  in  the  upper  part  are  explained  by  other  en- 
vironmental factors  unfavorable  for  growth  and  cone  pro- 
duction. In  the  Japanese  stone  pine  scrub  zone,  the  scrub 
height  is  related  to  winter  wind  strength  and  snow  depth 
of  their  habitats  (Okitsu  1987).  In  high-moimtain  areas, 
the  Japanese  stone  pine  needs  to  be  protected  by  snow- 
pack  to  evade  physical  damages  or  dehydration  caused  by 
strong  winds.  Short  scrub  height  indicates  strong  winds 
and  shallow  snow  cover  at  the  habitats  during  the  winter 
season,  while  tall  scrub  height  indicates  moderate  winds 
and  heavy  snowpack. 

On  Movmt  Kisokomagatake,  the  depth  of  snow  varies 
from  10  cm  to  3  m  according  to  landforms  and  wind  veloc- 
ity (Koiziuni  1974).  Especially  around  the  simimit,  most 
of  the  snow  is  blown  away  from  the  windward  slope  by 
strong  westerly  winds  (fig.  6).  Snow  accumulates  on  the 
leeward  side  of  the  ridges  or  rocks.  On  the  windward 
slope,  characterized  by  shallow  snow  cover,  the  Japanese 
stone  pine  grow  close  to  the  ground,  thus  protecting  them- 
selves from  strong  winds  (fig.  7).  Therefore,  vertical 
growth  of  the  Japanese  stone  pine  scrubs  is  disturbed  by 
winter  winds. 

Variation  in  snow  depth  caused  differences  in  the 
length  of  the  snow  melting  period.  According  to  3  years' 
observation  by  this  author,  snow  on  the  Japanese  stone 
pine  scrubs  located  on  windward  slopes  started  to  melt 


Figure  5 — Distribution  pattern  of  Japanese 
stone  pine  scrubs  in  the  lower  part  of  the 
Mount  Kisokomagatake  study  area.  They 
spread  extensively  and  cover  the  slopes  en- 
tirely from  the  ridges  down  to  the  forest  line. 


Table  1 — The  growth  and  pine  cone  production  in  each  part  of 
the  distribution  zone  viewed  from  different  altitude  of 
quadrats.  The  boundary  of  the  upper  and  the  lower  part 
is  2,850  m 


Cone  production  and  growth 

Upper  part 

Lower  part 

Cone  production  (pieces) 

4.1 

6.4 

Mean  scrub  height  (cm) 

36.2 

89.2 

Mean  stem  length  (cm) 

92.0 

180.0 

Mean  diameter  at  stem  base  (cm) 

2.2 

4.2 

Mean  annual  stem  elongation  (cm) 

2.4 

3.0 

Number  of  stems/m^ 

38.5 

20.0 

'The  values  indicate  the  means  of  20  plots  for  the  upper  part  and  12  plots 
for  the  lower  part. 


190 


Figure  6 — Mount  Kisokomagatake  summit  in 
early  spring  (1 991 ).  Snow  depth  varies  accord- 
ing to  wind  velocities  and  topography.  The 
western-most  part  (left)  where  snow  cover  was 
thin  due  to  strong  westerly  winds  shows  earliest 
snow  melt  in  spring. 


at  the  end  of  April.  On  the  leev^^ard  slopes,  the  scrubs 
became  completely  free  of  snow  by  the  middle  of  June. 
On  the  slopes  where  snow  remains  imtil  the  end  of  July, 
the  Japanese  stone  pine  is  replaced  by  alpine  snowbed 
vegetation.  On  the  other  hand,  shallow  snow  cover  causes 
early  snow  melt  and  solifluction  is  very  common.  Solifluc- 
tion  forms  stone  banked  terraces  and  cuts  off  the  roots  of 
young  plants.  For  this  reason,  the  Japanese  stone  pine 
scrubs  in  the  upper  part  are  not  able  to  cover  all  the 
slopes  and  thus  show  a  fragmented  distribution  pattern. 

These  severe  environmental  conditions  are  not  favor- 
able for  growth  or  cone  production  of  the  Japanese  stone 
pine.  The  main  cone  production  is  performed  by  large  and 


Figure  7 — Creeping  Japanese  stone  pine  in  the 
upper  part  of  the  study  area.  The  scrubs'  height  is 
10-15  cm.  The  stem  has  branches  flagged  toward 
the  leeward  side.  The  plot  shows  one  of  the  low- 
est cone  productions  (only  one  cone  for  1 0  stems) 
during  these  1 5  years. 


tall  scrubs  at  the  lower  elevations  where  winter  winds  are 
moderate  and  snow  is  evenly  distributed. 

ACKNOWLEDGMENTS 

I  thank  Professor  Dr.  F.-K.  Holtmeier,  Dr.  Wyman  C. 
Schmidt,  and  Dr.  W.  Schonenberger  for  their  invitation 
to  attend  the  workshop.  I  thank  Dr.  Professor  Hiroo 
Ohmori  and  Dr.  Takeei  Koizumi  for  their  valuable  advice 
in  carrying  out  this  study.  My  thanks  also  go  to  Dr.  Tad 
Weaver  and  Dr.  Penelope  Morgan  for  providing  me  with 
information  about  cone  scar  methods  on  whitebark  pine. 
Dr.  Mitsuhiro  Hayashida  offered  me  valuable  information 
in  preparation  for  the  workshop.  I  am  grateful  to  all  my 
colleagues  in  the  Geography  Department  of  Tokyo 
Gakugei  University,  who  enthusiastically  helped  me  in 
my  field  research  on  Mount  Kisokomagatake.  I  thank 
Mr.  Wataru,  Mrs.  Joyce  Mukai,  and  Mr.  Garrett  Vance 
who  checked  my  manuscript.  And  I  thank  Dr.  Miki 
Kobayashi  who  always  encourages  me  to  keep  studying 
Pinus  pumila. 

REFERENCES 

Hayashida,  M.  1989.  Seed  dispersal  and  regeneration  pat- 
terns of  Pinus  parviflora  var.  pentaphylla  on  Mt.  Apoi 
in  Hokkaido.  Research  Bulletins  of  the  College  Experi- 
mental Forests,  Faculty  of  Agriculture,  Hokkaido  Uni- 
versity. 46-1:  177-190. 

Kajimoto,  T.  1989.  Above-groimd  biomass  and  litterfall  of 
Pinus  pumila  scrubs  growing  on  the  Kiso  Mountain 
Range  in  central  Japan.  Ecological  Research.  4:  55-69. 

Koiziuni,  T.  1974.  Landschaftsokologische  Unter- 
suchungen  in  der  Alpine  Stufe  des  Kisokomagatake  in 
der  Japanischen  Zentralalpen  mit  Besonderer  Beriick- 
sichtigung  der  Vegeta  Strukturbodens.  Japanese  Jour- 
nal of  Ecology.  24(2):  78-91. 

Mizoguchi,  T.;  Okitsu,  S.  1987.  Cone  production  of  Pinus 
pumila  Regal  in  high  mountains  in  central  Japan  with 
special  attention  to  basal  diameter  and  annual  elonga- 
tion on  it.  In:  Transactions  of  the  98th  annual  meeting 
of  the  Japanese  Forestry  Society;  1987  April;  K5aishu: 
363-364. 

Morgan,  P.;  Bunting,  S.  C.  1992.  Technical  note:  cone 
crops  of  whitebark  pine.  Western  Journal  of  Applied 
Forestry.  7(3):  71-73. 

Nakashinden,  I.  1990.  Pine  cone  production  of  Pinus 
pumila  and  their  environment  in  Mt.  Kisokomagatake, 
Nagano  prefecture.  In:  Transactions  of  the  38th  autum- 
nal meeting  of  the  Association  of  Japanese  Geogra- 
phers; 1990  September;  Niigata:  104-105. 

Nakashinden,  I.  1991.  Pine  cone  production  of  Pinus 
pumila  reconstructed  from  pine  cone  scars.  In:  Transac- 
tions of  the  39th  spring  meeting  of  the  Association  of 
Japanese  Geographers;  1991  April;  Tokyo:  28-29. 

Okitsu,  S.  1984b.  Comparative  studies  on  the  Japanese 
alpine  zone,  with  special  reference  to  the  ecology  of 
Pinus  pumila  thickets.  Geographical  Review  of  Japan. 
57(11):  170-172. 

Okitsu,  S.  1987.  Pinus  pumila  zone.  In:  Ito,  K.,  comp.  Veg- 
etation of  Hokkaido.  Hokkaido,  Japan:  Hokkaido  Univer- 
sity Press:  129-167. 


191 


Okitsu,  S.;  Ito,  K.  1983.  Dynamic  ecology  of  the  Pinus 
pumila  community  of  Mts.  Taisetsu,  Hokkaido,  Japan. 
The  Environmental  Science,  Hokkaido  University.  6(1): 
151-184. 

Okitsu,  S.;  Mizoguchi,  T.  1986.  Cone  production  of  Pinus 
pumila  community  on  Mt.  Senjo,  Akaishi  Range,  cen- 
tral Japan.  In:  Transactions  of  the  97th  annual  meeting 
of  the  Japanese  Forestry  Society;  1986  April; 
Utsunomiya:  343-344. 

Okitsu,  S.;  Mizoguchi  T.  1990.  Relation  between  cone 
production  and  stem  diameter  and  stem  elongation 
of  Pinus  pumila  of  Japanese  high  mountains.  Japanese 
Journal  of  Ecology.  40(2):  49-55. 

Okitsu,  S.;  Mizoguchi,  T.  1991.  Cone  production  of  Pinus 
pumila  scrubs.  Japanese  Journal  of  Ecology.  41(2): 
101-107. 

Saito,  S.;  Kawabe,  M.;  Nakaoka,  T.  1989.  On  the  forest 
vegetation  of  Mt.  Upepesanke,  Tokachi,  Hokkaido.  (1): 
On  a  thicket  of  Pinus  pumila  at  the  ridge  of  1610  m. 


Bulletin  of  the  Higashi  Taisetsu  Museum  of  Natural 
History.  11:  23-34. 
Saito,  S.;  Kawabe,  M.  1990.  On  the  forest  vegetation  of 
Mt.  Higashi-Nupukaushinupuri,  Tokachi,  Hokkaido. 
(2):  On  two  thickets  of  Pinus  pumila.  Bulletin  of  the 
Higashi  Taisetsu  Museimi  of  Natural  History.  12: 
17-29. 

Tomback,  D.;  Linhart,  Y.  1990.  The  evolution  of  bird- 
dispersed  pines.  Evolutionary  Ecology.  4:  185-219. 

Weaver,  T.;  Forcella,  F.  1986.  Cone  production  in 
Pinus  albicaulis  forests.  In:  Shearer,  R.  C,  comp. 
Proceedings — conifer  tree  seed  in  the  Inland  Mountain 
West;  1985  August  5-6;  Missoula,  MT.  Gen.  Tech.  Rep. 
INT-203.  Ogden,  UT:  U.S.  Department  of  Agriculture, 
Forest  Service,  Intermountain  Research  Station:  68-76. 

Yanagimachi,  O.;  Ohmori,  H.  1991.  Ecological  status  of 
Pinus  pumila  scrub  and  the  lower  boimdary  of  the 
Japanese  alpine  zone.  Arctic  and  Alpine  Research. 
23(4):  242-435. 


192 


EFFECTS  OF  SEED  DISPERSAL  BY 
CLARK'S  NUTCRACKER  ON  EARLY 
POSTFIRE  REGENERATION  OF 
WHITEBARK  PINE 

Diana  F.  Tomback 


Abstract — Clark's  nutcrackers  {Nucifraga  columbiana)  bury 
whitebark  pine  (Pinus  alhicaulis)  seeds  in  recently  burned  ter- 
rain. Studies  of  early  postfire  whitebark  pine  regeneration  in  the 
Northern  Rocky  Mountains  indicate  that  seed  dispersal  by  nut- 
crackers (1)  may  be  more  effective  than  seed  dispersal  by  wind  in 
restocking  large  bums,  particularly  when  the  seed  source  is  unfa- 
vorably positioned  with  respect  to  prevailing  winds,  and  (2)  pro- 
duces a  negative  exponential  distribution  when  whitebark  pine 
regeneration  density  is  plotted  vs.  distance  from  the  seed  source. 
Whitebark  pine  seeds  can  remain  dormant  for  one  or  more  years 
after  dispersal,  which  results  in  continuity  of  regeneration  over 
time. 


1986),  although  updrafts  and  storms  may  disperse  some 
seeds  over  longer  distances.  Often,  the  regeneration  from 
nutcracker  caches  is  in  the  form  of  whitebark  pine  seed- 
Ung  clusters  (for  example  see  Tomback  1982,  1986;  Tomback 
and  others  1990, 1993),  which  may  result  in  mature  multi- 
genet  tree  clusters  (see  Tomback  and  Schuster,  these  pro- 
ceedings, and  references  therein).  In  this  paper  I  summa- 
rize the  results  of  recent  studies  that  compare  patterns  of 
early  postfire  regeneration  of  the  bird-dispersed  white- 
bark pine  and  its  wind-dispersed  forest  associates. 

REGENERATION  IN  WESTERN 
MONTANA 

Studies  of  two  large,  severe  subalpine  bums  in  western 
Montana  (Bitterroot  National  Forest,  Ravalli  County) 
illustrate  how  seed  dispersal  by  nutcrackers  affects  pat- 
terns of  early  postfire  regeneration  of  whitebark  pine  and 
associated  conifers.  These  two  study  areas  were  selected 
for  similarities  in  topography,  severity  of  bum,  forest 
types,  and  time  elapsed  since  fire.  In  this  region  of  west- 
em  Montana,  whitebark  pine  is  abundant  between  about 
2,290  and  2,620  m  elevation  (Amo  and  Hoff  1989;  Pfister 
and  others  1977),  and  the  prevailing  winds  in  late  sum- 
mer and  fall  are  from  the  west  and  southwest  (Finklin 
1983).  The  prefire  forest  in  both  study  areas  was  a  serai 
whitebark  pine  community.  Reported  results  fi-om  these 
studies  are  summarized  here  from  Tomback  and  others 
(1990,  1993). 

The  Sleeping  Child  Bum  resulted  from  a  wildfire  in 
1961  that  bumed  about  11,350  ha  of  forest  on  the  west 
slope  of  the  Sapphire  Range.  The  nearest  whitebark  pine 
seed  sovu*ce  is  mature  forest  above  2,250  m  on  the  eastern 
edge  of  the  bum.  Subalpine  fir,  Engelmann  spmce  (Picea 
engelmannii),  and  lodgepole  pine  (Pinus  contorta)  are  also 
present  in  this  upper  seed  source  as  well  as  around  the 
lower  perimeters  of  the  bum.  Because  the  prevailing 
winds  are  fi*om  the  west,  vdnd-dispersed  seeds  are  irregu- 
larly dispersed  into  the  eastem  portion  of  the  bum  fi'om 
the  upper  seed  source. 

In  1987,  we  established  a  3.6-km-long  transect  along 
an  eastwest  trending  ridge,  running  from  the  seed  soiirce 
into  the  northem  center  of  the  burn  (fig.  2  in  Tomback 
and  others  1990).  Elevation  along  the  ridge  decreases 
from  about  2,480  to  2,150  m  with  increasing  distance 
from  the  eastern  edge  of  the  bum.  Every  150  m  along  the 
transect  we  established  parallel  quadrats,  each  50  m  long, 
on  the  ridgetop,  north,  and  south  aspects  of  the  ridge  for 


Fire  is  an  ecologically  important,  recurring  event  in 
most  western  North  American  montane  forests,  including 
those  regions  where  whitebark  pine  {Pinus  albicaulis)  is 
found.  Fire-prone  areas  are  those  with  fuel  accumulation, 
dry,  windy  weather,  and  periods  of  frequent  lightning 
storms.  In  the  Northem  Rocky  Mountains,  fire  burns 
whitebark  pine  communities  every  50  to  300  years  (Amo 
1980;  Amo  and  Petersen  1983;  Romme  1980).  Stand- 
replacing  fires  result  in  the  renewal  of  successional  white- 
bark pine  communities,  and  underbums  kill  the  less  fire- 
resistant  subalpine  fir  {Abies  lasiocarpa).  In  the  absence 
of  fire  in  these  commimities,  whitebark  pine  is  often  re- 
placed by  more  shade-tolerant  conifers,  particularly  sub- 
alpine fir  (Amo  1986;  Amo  and  Hoff  1989). 

Studies  of  early  postfire  regeneration,  defined  here  as 
the  first  3  decades  after  fire,  provide  information  on  the 
dynamics  of  seed  dispersal  and  seedling  establishment  for 
whitebark  pine  and  associated  conifers.  Clark's  nutcrack- 
ers {Nucifraga  columbiana)  are  the  principal  dispersal 
agents  for  whitebark  pine  seeds,  in  contrast  to  other  coni- 
fers whose  seeds  are  wind-dispersed.  Nutcrackers  bury 
caches  of  one  to  15  or  more  ripe  whitebark  pine  seeds  at 
distances  from  parent  trees  of  a  few  meters  to  22  km 
(Hutchins  and  Lanner  1982;  Tomback  1978, 1982;  Vander 
Wall  and  Balda  1977;  see  review  in  Tomback  and  Linhart 
1990).  In  contrast,  most  vraid-dispersed  seeds  land  only 
within  about  120  m  of  parent  trees  (McCaughey  and  others 


Paper  presented  at  the  International  Workshop  on  Subalpine  Stone 
Pines  and  Their  Environment:  The  Status  of  Our  Knowledge,  St.  Moritz, 
Switzerland,  September  5-11,  1992. 

Diana  F.  Tomback  is  Associate  Professor  of  Ecology,  Department  of 
Biology  and  Center  for  Environmental  Sciences,  University  of  Colorado 
at  Denver,  Denver,  CO  80217-3364. 


193 


a  total  of  63  quadrats.  Quadrat  widths  varied  in  preset 
increments  from  1.25  to  15  m,  with  respect  to  local  densi- 
ties of  whitebark  pine.  For  each  quadrat  we  recorded  all 
conifer  regeneration. 

The  Saddle  Moimtain  Burn  resulted  from  a  wildfire  in 
1960  that  spread  over  1,240  ha  of  forest  in  the  Bitterroot 
Mountains  west  of  Lost  Trail  Pass.  Running  northeast 
to  southwest  in  the  longest  dimension,  the  burned  area 
ranges  in  elevation  from  about  1,950  to  2,475  m.  On  the 
southwest  edge  of  the  burn  above  2,275  m,  the  whitebark 
pine  seed  source  is  positioned  so  that  the  burned  area  is 
leeward  with  respect  to  prevailing  winds.  Subalpine  fir, 
Engelmann  spruce,  and  lodgepole  pine  also  occur  in  this 
upper  seed  source  and,  with  Douglas-fir  {Pseudotsuga 
menziesii),  around  the  lower  perimeters  of  the  burn. 

In  1988  we  established  two  parallel  3.2-km-long  tran- 
sects beginning  at  the  seed  source  and  heading  northeast 
through  the  southern  portion  of  the  burn  (fig.  4  in  Tomback 
and  others  1990).  We  set  up  41  quadrats  and  recorded  all 
regeneration  according  to  the  general  methodology  out- 
lined earlier  for  the  Sleeping  Child  Burn  (for  details  of 
methodology,  see  Tomback  and  others  1990,  1993). 

Mean  whitebark  pine  regeneration  density  sampled  per 
quadrat  in  the  Sleeping  Child  Burn  was  much  greater 
than  that  of  its  shade-tolerant  competitors,  subalpine  fir 


Table  1 — Mean  density  and  standard  deviation  (S.D.)  per  quadrat 
for  conifer^  regeneration  26  years  and  28  years  after  fire 
in  tlie  Sleeping  Child  and  Saddle  Mountain  Burns, 
respectively.  Based  on  Tomback  and  others  (1 990) 

Regeneration  sites  per 
ABLA      PIEN       PICO  PSME 


PIAL 


Sleeping  Child 

X  0.070  0.008  0.004  0.039 

S.D.  .104  .014  .009  .066 

Saddle  Mountain 

X  .044  .046  .030  .323 

S.D.  .050  .068  .044  .355 


0.019 
.032 


TIAL  =  whitebark  pine,  ABLA  =  subalpine  fir,  PIEN 
PSME  =  Douglas-fir. 


Engelmann  spruce. 


and  Engelmann  spruce,  and  greater  than  the  shade- 
intolerant  lodgepole  pine  as  well  (table  1).  Scatterplots  of 
quadrat  densities  vs.  distance  from  the  upper  seed  source 
revealed  interesting  differences  in  regeneration  patterns 
among  the  conifers  (fig.  1).  The  scatterplot  for  whitebark 
pine  was  a  negative  exponential  curve,  with  the  highest 
densities  near  the  seed  source  and  a  long  tail  indicating 


0.512 


0.400  - 


0.200  - 

0 


'3) 
c 

Q 


p.  albicaulis 


•  •    •  •  • 
•  •     •  •  • 

.•    •  .  J  J  I  It         ••  •  • 


0.296  |-      p.  contorta 


0.200 


0.100 


•      •  •  • 

-  •  •      •  • 


•     •  • 


0.08  r       A.  lasiocarpa 


0.05 


0.02 


0.056  r      p  engelmannii 


0.040 


•  •  • 


0.020 


•  I 


•  I  • 
•  •  • 


I  I  »» w    I  l,«  «  i  »»  .  I  »  y  I  «  I  I 


1000 


2000 


3000  3650 


1000 


2000 


3000  3650 


Distance  (m) 


Figure  1 — Scatterplots  of  regeneration  density  per  quadrat  versus  distance  from  the  whitebark  pine 
seed  source  for  four  conifer  species  in  the  Sleeping  Child  Burn.  Clockwise  from  left:  whitebark  pine, 
lodgepole  pine,  Engelmann  spruce,  subalpine  fir.  From  Tomback  and  others  (1990). 


194 


some  regeneration  beyond  3.6  km.  Quadrat  density  vs. 
distance  in  subalpine  fir  showed  a  negative  linear  rela- 
tionship, indicating  that  the  primary  seed  soiirce  for  this 
species  was  also  the  upper,  unbumed  forest.  In  contrast, 
the  scatterplot  for  Engelmann  spruce  was  nearly  flat, 
with  the  highest  values  at  the  lower  end  of  the  transect, 
suggesting  that  seeds  may  be  blown  into  the  ridge  area 
from  both  the  higher  and  lower  seed  sources.  For  lodge- 
pole  pine,  densities  increased  with  distance  from  the  up- 
per, iinbumed  forest  (are  higher  at  lower  elevations). 
Before  the  fire,  the  lower  and  middle  elevations  of  this 
area  were  dominated  by  lodgepole  pine,  and  thus  much  of 
the  postfire  regeneration  may  have  come  from  seeds  from 
serotinous  cones  within  the  bum  (Lotan  1976).  With  the 
exception  of  Engelmann  spruce,  the  density  vs.  distance 
relationships  were  statistically  significant  atP  <  0.01  or 
greater  (Tomback  and  others  1990). 

For  the  Saddle  Moimtain  Burn,  mean  whitebark  pine 
regeneration  density  was  not  significantly  different  than 
that  for  the  Sleeping  Child  Bum,  although  somewhat 
lower  (table  1).  The  density  was  actually  comparable  to 
that  obtained  for  the  ridgetop  quadrats  (west  facing)  in 
the  Sleeping  Child  Bum.  In  contrast,  the  densities  for 
subalpine  fir  and  Engelmann  spruce  were  much  greater 
than  those  in  the  Sleeping  Child  Burn.  Lodgepole  pine 
was  particularly  abimdant  as  well.  Scatterplots  of  quad- 
rat density  vs.  distance  from  the  upper  seed  source  were 
comparable  to  those  from  Sleeping  Child  (fig.  2). 


0.14 


0.10 


r  •• 


(0 

c 

0) 
Q 


0.05 

0 

0.30 
0.20 

0.10 


P.  albicaulis 


•      •  • 


A.  lasiocarpa 


••  • 


•  •  •  • 

•  ■  •  •   •  • 

•  ,  •      ••••••  • 


1000 


2000 


3000 


Distance  (m) 

Figure  2 — Scatterplots  of  regeneration  density 
per  quadrat  versus  distance  from  the  whitebark 
pine  seed  source  for  whitebark  pine  (top)  and 
subalpine  fir  (bottom)  in  the  Saddle  Mountain 
Burn.  From  Tomback  and  others  (1990). 


Table  2 — Maximum  age  differences  within  clusters  of  whitebark 

pine  seedlings.  Cluster  sample  size  for  the  Sleeping  Child 
Burn  =  87;  for  the  Saddle  Mountain  Burn  =  7.  Based  on 
table  4,  Tomback  and  others  (1993) 


Maximum  age 
differences 

Sleeping  Child 

Saddle  Mountain 

Years 

n 

Percent 

n 

Percent 

0 

30 

34.5 

1 

14.3 

1 

24 

27.6 

4 

57.1 

2 

12 

13.8 

1 

14.3 

3 

10 

11.5 

0 

0 

4 

6 

6.9 

1 

14.3 

5 

1 

1.2 

6 

1 

1.2 

7 

2 

2.3 

8 

1 

1.2 

Whitebark  pine  regeneration  showed  a  negative  exponen- 
tial distribution,  and  subalpine  fir  a  similar  but  more  lin- 
ear negative  relationship  between  quadrat  density  and 
distance  from  the  upper,  unbiu-ned  forest  (density  vs. 
distance  relationships  significant  at  P  <  0.001  for  both). 
Similar  to  findings  for  the  Sleeping  Child  Bum,  the  En- 
gelmann spruce  scatterplot  was  nearly  flat.  For  both 
Douglas-fir  and  lodgepole  pine,  density  increased  with 
distance  from  the  upper,  imbumed  forest;  major  seed 
sources  for  both  species  were  at  the  lower  edges  of  the 
bum  and,  in  the  case  of  lodgepole  pine,  seeds  from  within 
the  bum.  However,  the  relationship  between  density  and 
distance  was  not  statistically  significant  for  these  latter 
three  conifers. 

The  distinctive  negative  exponential  shape  of  the  white- 
bark pine  regeneration  curve  probably  results  from  two 
factors:  (1)  whitebark  pine  seeds  are  only  available  at  the 
upper  elevations  of  these  burns,  and  (2)  nutcrackers  tend 
to  bury  higher  densities  of  seeds  near  parent  trees.  Ap- 
parently, nutcrackers  do  cache  seeds  at  long  distances 
from  parent  whitebark  pine  trees,  because  we  found  low 
densities  of  regeneration  up  to  8  km  from  the  whitebark 
pine  seed  source  (Tomback  and  others  1990).  It  could  be 
argued  that  the  concomitant  decrease  in  elevation  in  both 
study  areas  was  a  confounding  variable  in  the  whitebark 
pine  density  vs.  distance  relationship.  However,  recently, 
I  tested  this  possibility  by  examining  whitebark  pine  re- 
generation along  a  3.4-km  transect  following  an  eleva- 
tional  isocline  in  the  Sundance  Bum  of  northern  Idaho. 
Like  the  Sleeping  Child  and  Saddle  Mountain  Bxirns,  the 
whitebark  pine  seed  source  was  restricted  to  one  side  of 
the  bum,  and  the  resulting  regeneration  curve  followed 
a  negative  exponential  distribution  (Tomback  1992). 

It  appears  that  seed  dispersal  by  nutcrackers  has 
provided  whitebark  pine  an  advantage  over  its  shade- 
tolerant  competitor,  subalpine  fir,  in  the  upper  subalpine 
of  the  Sleeping  Child  Burn.  Because  wind-dispersed  fir 
seeds  from  the  upper  seed  source  must  travel  against  the 
prevailing  winds  and  cover  a  long  distance,  the  regenera- 
tion density  of  fir  is  lower  than  that  for  whitebark  pine. 
In  contrast,  in  the  smaller  Saddle  Mountain  Burn,  white- 
bark pine  and  subalpine  fir  have  comparable  regeneration 


195 


densities.  In  this  situation,  dispersal  of  subalpine  fir  seeds 
(and  Engelmann  spruce  seeds)  is  favored  both  by  the  pre- 
vailing winds  from  the  west  and  the  much  smaller  size  of 
the  burn.  Thus,  for  restocking  burns,  seed  dispersal  by 
nutcrackers  may  provide  whitebark  pine  an  advantage 
over  wind-dispersed  conifers  when  (1)  the  size  of  the  burn 
is  large  compared  to  distances  traveled  by  wind-blown 
seeds,  (2)  the  seed  source  is  unfavorably  positioned  with 
respect  to  prevailing  winds,  and  (3)  the  seed  source  is  1  km 
or  more  from  the  perimeter  of  the  burn. 

Another  consequence  of  seed  dispersal  by  nutcrackers 
is  that  seedling  or  sapling  clusters  occurred  at  more  than 
40  percent  of  the  whitebark  pine  regeneration  sites  in  both 
the  Sleeping  Child  and  Saddle  Mountain  Burns  (Tomback 
and  others  1990,  1993).  Cluster  sizes  in  the  Sleeping 
Child  Burn  ranged  from  two  to  eight  stems  with  an  over- 
all mean  of  1.9  stems  per  regeneration  site  (n  =  455  total 
sites).  In  the  Saddle  Mountain  Burn,  clusters  ranged 
from  two  to  10  stems,  with  an  overall  mean  of  2.0  stems 
per  regeneration  site  (n  =  164  sites).  We  aged  two  or  more 
cluster  members  from  a  cluster  subsample  from  each  of 
the  study  areas  (Tomback  and  others  1993).  For  the 
Sleeping  Child  Burn,  there  were  no  differences  in  age 
among  seedlings  within  a  cluster  for  34  percent  of  the 
clusters,  differences  of  1  year  for  28  percent  of  the  clus- 
ters, and  differences  of  2  to  8  years  within  clusters  for  the 
rest  of  the  subsample  (total  n  =  87  clusters;  all  seedlings 
aged  in  63  clusters)  (table  2).  Of  the  small  number  of 
clusters  sampled  in  the  Saddle  Mountain  Burn,  five  clus- 
ters (71  percent)  had  age  differences  of  0  to  1  year,  and 
the  remaining  two  clusters  had  seedlings  differing  by 
2  and  4  years,  respectively  (table  2).  These  resiilts  strongly 
suggest  that  asynchronous  germination  of  seeds  may  oc- 
cur in  some  clusters.  This  could  be  explained  by  delayed 
maturation  of  some  seeds  or  different  sensitivities  to  envi- 
ronmental conditions.  Laboratory  (Leadem  1986;  Pitel  and 
Wang  1990)  and  experimental  field  studies  (McCaughey 
1993)  indicate  that  embryo  underdevelopment  and  de- 
layed germination  are  tjrpical  in  whitebark  pine  seeds. 
Delayed  germination  of  seed  caches  would  result  in  seed- 
ling establishment  following  little  or  no  cone  production 
and,  thus,  continuity  of  regeneration  over  time. 

We  compared  age  structiire  (year  of  tree  establishment 
vs.  time  since  fire)  of  regeneration  for  whitebark  pine 
and  subalpine  fir  in  both  the  Sleeping  Child  and  Saddle 
Mountain  Burns  to  determine  any  important  differences 
(Tomback  and  others  1993).  Both  species  appeared  to  be 
similarly  affected  by  environmental  and  ecological  factors: 
a  disproportionately  high  percentage  of  regeneration  was 
established  between  1977  and  1985,  about  17  to  25  years 
after  fire  (see  fig.  1  in  Tomback  and  others  1993).  Thus, 
major  tree  recruitment  was  episodic.  A  small  increase 
in  mean  January  through  August  precipitation  be